Research on detection method to sudden contaminant of underwater closed cabin based on chi-square distribution

To determine whether there is sudden contaminant in the closed-cabin, we can judge on a basis that whether the corresponding concentration of contaminants in the cabin higher than a certain threshold value. However, as noise impact of contaminants measurement instruments, it is difficult to determine accurately whether the sudden change of concentration is due to strong noise of monitoring instruments or caused by a sudden contaminant source, which results in the delay and inaccuracy of contaminant source identification and concentration prediction. In order to solve this problem, a new method to detect sudden contaminant based on Χ2 distribution is presented in this paper. The detection method could overcome the impact of sensor noise, and get the information of sudden contaminant starting time accurately, which could ensure the reliability of the contaminant concentration prediction and hazard warning. Simulation results show the effectiveness of the algorithm.