Fault detection and diagnosis in activated effluent disposal process based on PCA

As one of approaches of sewage disposal, activated sludge method is the most well-known technique in present world that has been used primarily in industrial organic sewage disposal and municipal sewage secondary disposal. Considering complexity and particularity of the process, this paper proposes a method named PCA(PCA-Principal component analysis), which can be used to eliminate nonlinear correlation between every two variates and reduce noise influence as well as detect and diagnose faults in activated sewage disposal process by using T² statistic, Q statistic, the judgment on threshold as well as contribution charts. It is shown that PCA method can be applied to sewage disposal which is a multi-variate monitoring process.