### 1. Introduction

^{3}per day with the total facility capacity of 24,735,000 m

^{3}per day. Also, there were 2,332 sewage treatment plants treating below 500 m

^{3}per day with the total facility capacity of 171,000 m

^{3}per day; and consequently, the percent of population being supplied with sewage was 89.4%, which was not lacking compared to other developed countries in the world [1]. Also, a vast improvement has been made in terms of sewage treatment processes compared to the days when there was only the activated sludge process, which targeted organic matters; now, there are advanced sewage treatment processes that target various nutrients and poisonous substances. However, there is still huge room for improvement in terms of the operation and management technology of sewage treatment plants.

### 2. Materials and Methods

### 2.1. Selected Sewage Treatment Plant

### 2.2. Multiple Regression Analysis

_{Mn}) and RO_TN (2nd settling tank effluent T-N) were set as dependent variables and the remaining 30 variables were set as independent variables (Table 1).

### 2.3. Analysis Method

*k*is the number of independent variables and

*n*is the number of samples.

*y*

*is the*

_{i}*i-*th observed value,

*ŷ*

*is the*

_{i}*i-*th predicted value, and

*l*is the total number of values in the prediction data.

### 3. Results and Discussion

_{Mn}) and RO_TN (2nd settling tank effluent T-N), all values resulting from the F-test on each case were below 0.05, which is statistically significant.

### 3.1. RO_COD (2nd Settling Tank Effluent COD_{Mn})

### 3.2. RO_TN (2nd Settling Tank Effluent T-N)

### 3.3. Selection of Multiple Regression Model

^{2}(coefficient of determination) value, accuracy of prediction, root mean squared error, and variance pattern’s predictabilities.

_{Mn}), case 4, which scored the highest in terms of R

^{2}(coefficient of determination) value and accuracy of prediction, was selected and a regression model could be constructed based on the regression coefficients of each independent variable (Table 5).

### 3.4. Reviewing the Influence of Operation Parameters

_{Mn}), the order of the highest standardized regression coefficient among 11 independent variables was OUTF, I_COD, AE_MLSS (Table 7).

_{Mn}), the order of highest standardized regression coefficient among the 8 independent variables was AE_TEMP, I_TN, FM (Table 8).

### 3.5. Reviewing the Economic Operation Condition of the Bioreactor

_{Mn}and TN in terms of energy savings. In order to provide the economic operation condition, first, the power consumption of the equipment (blowers, internal return sludge pumps, excess sludge pumps, etc.) for operating the bioreactor should be reviewed. The economic operation condition can be provided by considering the effluent quality standards and the amount of power consumption (Fig. 6).

### 4. Conclusions

_{Mn}and T-N based on multiple regression analysis, it was shown that the accuracy of prediction came out to be above 0.93 and 0.84, respectively. Also, through this model, the variance pattern of the actual values was proven to have been predicted fairly well.

_{Mn}), case 4 was selected, because it scored the highest in terms of R

^{2}(coefficient of determination) value and accuracy of prediction; and in the case of dependent variable RO_TN (2nd settling tank effluent T-N), case 1, which scored the highest in terms of root mean squared error and accuracy of prediction, was selected.

_{Mn}was AE_MLSS and F/M for controlling T-N.

_{Mn}and T-N were verified by reviewing the standardized regression coefficients of the multiple regression models constructed in this study. Predicting the concentration of COD

_{Mn}and T-N in the 2nd settling tank effluent was possible by following the operation conditions using the selected multiple regression model. If the data on the energy spent on each operation parameter can be collected, then the operation parameter that conserves energy without violating the effluent quality standards of COD

_{Mn}and TN can be determined using the regression model and the standardized regression coefficients. These results can provide appropriate operation guidelines to conserve energy to the operator at a sewage treatment plant that consumes a lot of energy.