The present paper deals with treatment of slaughterhouse wastewater by conducting a laboratory scale sequencing batch reactor (SBR) with different input characterized samples, and the experimental results are explored for the formulation of feedforward backpropagation artificial neural network (ANN) to predict combined removal efficiency of chemical oxygen demand (COD) and ammonia nitrogen (
The sequencing batch reactor (SBR) is the most promising and viable of the proposed activated sludge modifications for the removal of organic carbon and nutrients [
Artificial neural networks (ANNs) are now used in many areas of science and engineering and considered as promising tool because of their simplicity towards simulation, prediction, and modeling [
ANNs are mathematical systems comprised of a number of “processing units” that are linked via weighted interconnections. The tool helps in mapping a set of input data into a corresponding set of output data after learning a series of past process data from a given system. Moreover, the neural network model possesses a distinctive ability of learning nonlinear functional relationship without requiring a structural knowledge of the process to be modeled. Several examples of successful development of software sensor mechanism based on the neural network approach are described in the literature [
The objective of the present study was application of feedforward backpropagation artificial neural network (ANN) modeling to predict the performance of a laboratory scale sequencing batch reactor operated under three different combinations of aerobic-anoxic sequential hours of cycles, namely, (4 hour aerobic + 4 hour anoxic), (5 hour aerobic + 3 hour anoxic), and (5 hour aerobic + 4 hour anoxic) of total react period in terms of COD removal and
Heterotrophic microbes were actively acclimatized in the laboratory environment by inoculating sludge collected from an aeration pond of a small-scale slaughterhouse located nearby to the institute. The mixture of sludge was continuously aerated with intermittent feeding with dextrose solution having concentrations of 1000 mg/L and ammonium chloride (NH4Cl) having concentration of 200 mg/L as a carbon and nitrogen source, respectively. The acclimatization process was continued for an overall period of 90 days. The biomass growth was monitored by the magnitude of sludge volume index (SVI) and mixed liquor suspended solid (MLVSS) concentration in the reactor. pH in the reactor was maintained in the range 6.8–7.5 by adding required amount of sodium carbonate (Na2CO3) and phosphate buffer. The seed acclimatization phase was considered to be over when a steady state condition was observed in terms of equilibrium COD and
Denitrifying seed was cultured separately in 2.0 L capacity aspirator bottle under anoxic condition. 500 gm of digested sludge obtained from a local sewage treatment plant (STP) was added to 1.0 L of distilled water. The solution was filtered and added to 500 mL of untreated slaughterhouse wastewater. The resulting solution was acclimatized for denitrification purpose using dextrose as carbon source and potassium nitrate (KNO3) as the source of nitrate nitrogen (
The experiment that was performed in a laboratory scale sequencing batch reactor of 20.0 L effective volume was fabricated by Plexiglass sheet of 6 mm thickness which is shown schematically in Figure
Experimental setup of SBR.
The treatability study of slaughterhouse wastewater in SBR was investigated for the wastewater sample collected from outlet of the slaughterhouse, as mentioned above. The raw and primary-treated wastewater was first characterized and subsequently treatability experiment was carried out in the laboratory scale SBR unit. The following analytical parameters, namely, pH, TSS, VSS, DO, BOD5 at 20°C, COD, TKN, and Combination-1: 4-hour aerobic react period and 4-hour anoxic react period; Combination-2: 5-hour aerobic react period and 3-hour anoxic react period; Combination-3: 5-hour aerobic react period and 4-hour anoxic react period.
In the ammoniacal nitrogen removal phase, the reactor was continuously aerated by a compressor of capacity 0.25 hp. during the fill period and for the entire aerobic react phase. The mechanical mixer was operated continuously with a speed of 400 rpm from the beginning of the fill phase to the end of the total react phase for proper mixing of liquid in the reactor. During the draw phase, the supernatant wastewater that was decanted until the liquid volume in the reactor decreased to 4.0 L. SRT (solid retention time) was manually controlled by withdrawal of volume of the mixed liquor from the reactor every day at the onset of the commencement of settle phase. The reactor was continuously run for 120 days. The treated volume of slaughterhouse wastewater was taken as 16 L. The initial pH values in the reactor were kept in between 7.0 and 8.0, whereas the sludge volume index (SVI) has been kept within the range of 75–85 mL/gm, for obtaining good settling property of the biomass. The SRT of 20–25 days is maintained for carbon oxidation and nitrification in the present SBR system for treatment of wastewater as suggested by Tremblay et al. [
During the time course of the study, 100 mL of sample was collected from the outlet of the reactor at every 1.0 hour interval, after completion of the fill period. The samples were analyzed for the following parameters, namely, pH, DO, MLSS, MLVSS, COD,
The architecture of ANN consists of input layer, one or more hidden layer, and output layer. Each layer of the network consists of a number of interconnected processing elements called as neurons. These neurons interact with each other with the help of the weight. Each neuron is connected to all the neurons in the next layer. The data is presented to the neural network in the input layer. The output of the neural network is presented by output layer for the given input data. The hidden layers enable these networks to compute complicated relations between input and output. The structure of feedforward neural network is exhibited in Figure
Architecture of ANN Model.
The number of hidden layers is to be selected in respect of the complexity of the problem. Generally one hidden layer is sufficient for investigation of most of the problem. The number of neurons in the hidden layer is selected by trial and error method starting from minimum value and then to be increased depending on the nature of problem. The training of neural network is carried out by presenting a series of input data and target output values. The parameters affecting the output should be selected as input parameters. The backpropagation training algorithm has been widely used to model various problems in environmental engineering. In the backpropagation training algorithm, the neurons in the hidden layer and output layer process its input by multiplying each input by its weight, summing the product, and then processing the sum using a nonlinear transfer function, also called activation function, to produce results. The most common transfer function used is the sigmoid function. The learning in the neural network takes place by modifying weights of the neurons according to the error between the actual output value and the target output values. The changes in weight are proportional to negative of the derivative of the error.
The backpropagation is essentially a gradient descending method to minimize the network error function equation:
Before starting training of an ANN, weights are initially randomized. Based on the error propagation, weights are adjusted based on
The training of ANN model was carried out by presenting the complete input data set to the network and continued till the average MSE is minimized. After the training is over, the trained neural network deemed to be reproduced by the target output values for provided training data. Weights of the trained neurons are then stored in the neural network memory. Testing of the trained network is carried out by presenting the set of test data and then comparing the output of the network with the actual values of the output. The performance of formulated ANN model can be measured by several statistical parameters, such as coefficient of determination
The numbers of input and output neurons are fixed according to the nature of the problem. In the present study, only one hidden layer was selected. The number of neurons in the hidden layer was selected from 2 to 30. The inputs to the neural network include the amount of MLVSS (mg/L), initial concentration of COD (mg/L), initial concentration of
The input and output variables in the present study consist of different characteristics and importance level, resulting into varying response to the neural network. The ANN model training would be more efficient if preprocessing steps are to be performed on the input and target data, by making the preprocessing exercise useable in real applications. The input parameters were scaled up in the range 0.2 to 0.8. The preprocessing of the data could be performed by the algorithm as given in
After preprocessing of the training data set, the new input sets are fed into the trained networks with minimum and maximum vectors for the training set. In order to compare the neural network results with the observed results, the rescaled outputs need to be converted back within the same range for the original targets. The algorithm as given in (
A three-layered feedforward neural network with backpropagation training algorithm was used to validate the ANN model. The tangent-sigmoid transfer functions (tansig) in between input and hidden layer and a linear transfer function (purelin) in between hidden layer and output layer were used. The Levenberg-Marquardt algorithm was used for ANN model training. During training process, small weights were assigned to the connection between neurons in a random way. Weights were modified until the error between the predicted and experimental values of adsorption efficiency are minimized. It is desired that the difference between predicted and observed values should be as small as possible. During testing process, the network was tested for its generalization ability with the observed output after the training process was completed. When the neural networks are tested successfully, they can be used for prediction.
The feedforward backpropagation (BP) algorithm with Levenberg-Marquardt training was applied for neural network model development. The BP is an approximate steepest descent algorithm where an MSE is used as performance function. In the neural network development, different number of hidden layers, the number of neurons in each layer, and the type of transfer function for each neuron were analyzed all with a learning rate of 1.0 [
In order to measure the pollutant removal efficiency and the performance of three neural network models developed for biological system, different types of statistical parameters were used to estimate the error. In the present work, ME, MSE, and RMSE were selected to measure the network performance of Models “A,” “B,” and “C” under three different combinations of aerobic-anoxic sequence, namely,
The optimal architecture of the ANN model and its parameter variation were determined based on the minimum value of the MSE of the training and prediction set as suggested by Ekici and Aksoy [
The slaughterhouse wastewater samples were collected from two different locations: (i) the raw (untreated) wastewater from the main collection pit and (ii) the primary treated effluent from the inlet box of aeration basin. The wastewater samples were collected 6 (six) times over the entire course of the study in 25.0 L plastic containers and stored in a refrigerator at approximately 4.0°C. The above wastewater samples (raw and primary treated) were characterized in the laboratory with respect to the parameters as exhibited in Table
Characteristics and composition of slaughterhouse wastewater.
Parameters |
Raw wastewater range | Pre treated wastewater range |
---|---|---|
pH | 8.0–8.5 | 7.5–8.5 |
Total suspended solids | 10120–14225 | 4255–5340 |
Total dissolved solids | 6345–7840 | 2800–3830 |
COD | 4185–5240 | 1930–2096 |
BOD5 at 20°C | 3000–3500 | 1010–1465 |
Total Kjeldahl nitrogen (TKN) | 450–560 | 315–412 |
|
215–310 | 110–130 |
The performance of the SBR system in the slaughterhouse wastewater treatment for combined carbon and nitrogen removal has been plotted in Figures
Carbon oxidation, nitrification, and denitrification profiles for slaughterhouse wastewater treatment in
Carbon oxidation, nitrification, and denitrification profiles for slaughterhouse wastewater treatment in
Carbon oxidation, nitrification, and denitrification profiles for slaughterhouse wastewater treatment in
At the end of anoxic react period, under
Three models were developed for combined carbon oxidation and nitrification of slaughterhouse wastewater in the following combinations: Model “A” (Combination-1): 4-hour aerobic and 4-hour anoxic react period in SBR system; Model “B” (Combination-2): 5-hour aerobic and 3-hour anoxic react period in SBR system; Model “C” (Combination-3): 5-hour aerobic and 4-hour anoxic react period in SBR system.
The ANN model was developed for biological removal of combined organic carbon and nitrogen of a slaughterhouse wastewater in a sequencing batch reactor using experimental data. Table
Range of variables in ANN study.
Serial number | Variables | Ranges |
---|---|---|
1 | Initial MLVSS concentration (mg/L) | 1810–2600 |
2 | Initial concentration of COD (mg/L) | 1910–2096 |
3 | Initial concentration of |
112–130 |
4 | pH | 7.0-8.0 |
5 | Initial DO level (mg/L) | 2.4–4.5 |
6 | Total react time (hour) | 8-9 |
7 | COD/ |
0–100 |
Data set for ANN Model “A” under (
Serial number | Initial COD |
Initial |
Initial MLVSS |
Time |
pH | Initial DO |
% COD |
% |
---|---|---|---|---|---|---|---|---|
1 | 2004.25 | 112.41 | 2155.41 | 8 | 7.1 | 3.8 | 86.45 | 86.92 |
2 | 2077.11 | 118.55 | 1844.22 | 8 | 7.4 | 2.4 | 88.11 | 88.52 |
3 | 1922.64 | 117.34 | 1863.57 | 8 | 7.1 | 3.5 | 89.62 | 88.81 |
4 | 1913.21 | 121.46 | 2482.36 | 8 | 7.0 | 2.7 | 85.37 | 84.44 |
5 | 2030.25 | 117.52 | 2272.21 | 8 | 7.2 | 3.6 | 89.33 | 85.83 |
6 | 1938.48 | 120.25 | 2212.54 | 8 | 7.0 | 3.7 | 83.21 | 85.11 |
7 | 1925.41 | 124.36 | 2475.45 | 8 | 8.0 | 3.8 | 80.19 | 84.9 |
8 | 2024.17 | 122.24 | 1940.26 | 8 | 7.0 | 4.4 | 86.08 | 85.25 |
9 | 2094.40 | 113.84 | 2131.22 | 8 | 7.5 | 4.0 | 88.25 | 86.54 |
10 | 2040.33 | 116.55 | 2075.35 | 8 | 7.0 | 2.8 | 88.65 | 82.97 |
11 | 1998.34 | 125.52 | 2322.14 | 8 | 7.0 | 3.0 | 85.22 | 88.54 |
12 | 1917.24 | 127.84 | 2391.22 | 8 | 7.2 | 3.0 | 85.39 | 89.31 |
13 | 1989.17 | 128.66 | 1917.27 | 8 | 7.5 | 3.6 | 84.11 | 82.58 |
14 | 1997.57 | 129.88 | 1968.26 | 8 | 8.0 | 3.5 | 86.05 | 90.31 |
15 | 2085.27 | 124.24 | 2012.55 | 8 | 7.0 | 3.0 | 90.42 | 84.22 |
16 | 2054.22 | 122.26 | 2048.22 | 8 | 7.4 | 3.0 | 91.94 | 86.27 |
17 | 2014.11 | 122.31 | 2144.24 | 8 | 7.2 | 3.7 | 84.18 | 88.55 |
18 | 1984.27 | 116.27 | 1950.25 | 8 | 7.1 | 3.8 | 84.62 | 89.33 |
19 | 1932.92 | 127.57 | 1995.27 | 8 | 7.5 | 4.3 | 85.21 | 86.28 |
20 | 2015.28 | 123.94 | 2212.51 | 8 | 7.5 | 4.4 | 85.22 | 82.85 |
21 | 2008.15 | 122.27 | 2352.81 | 8 | 7.0 | 4.2 | 83.64 | 81.22 |
22 | 2017.47 | 126.67 | 2432.55 | 8 | 7.6 | 4.0 | 88.52 | 81.6 |
23 | 2037.81 | 120.76 | 2245.28 | 8 | 7.6 | 4.2 | 91.39 | 82.27 |
24 | 1938.58 | 119.41 | 2185.35 | 8 | 7.4 | 3.6 | 85.35 | 85.22 |
25 | 2096.55 | 125.22 | 2265.21 | 8 | 7.5 | 3.5 | 84.29 | 85.52 |
26 | 2036.28 | 115.48 | 2258.36 | 8 | 7.1 | 3.4 | 86.55 | 87.22 |
27 | 1965.25 | 114.88 | 2424.37 | 8 | 7.4 | 4.0 | 87.56 | 84.25 |
28 | 1975.27 | 126.48 | 2575.36 | 8 | 7.0 | 2.9 | 86.34 | 87.54 |
29 | 2072.55 | 128.85 | 1850.29 | 8 | 7.5 | 3.0 | 90.22 | 88.25 |
Data set for ANN Model “B” under (
Serial number | Initial COD (mg/L) | Initial |
Initial MLVSS |
Time |
pH | Initial DO |
% COD |
% |
---|---|---|---|---|---|---|---|---|
1 | 2044.23 | 113.01 | 2055.46 | 8 | 7.2 | 2.8 | 84.42 | 88.90 |
2 | 2057.15 | 119.56 | 1944.22 | 8 | 7.5 | 3.4 | 84.16 | 86.55 |
3 | 1972.65 | 118.32 | 1963.51 | 8 | 7.0 | 3.2 | 86.62 | 89.88 |
4 | 1943.22 | 120.44 | 2382.33 | 8 | 7.5 | 2.9 | 82.35 | 86.45 |
5 | 2050.22 | 116.50 | 2172.24 | 8 | 7.4 | 3.1 | 88.34 | 89.85 |
6 | 1918.08 | 130.24 | 2412.52 | 8 | 7.5 | 3.5 | 84.26 | 87.15 |
7 | 1955.47 | 121.35 | 2575.44 | 8 | 8.0 | 3.8 | 82.39 | 86.96 |
8 | 2044.12 | 123.28 | 1940.22 | 8 | 7.5 | 4.0 | 85.58 | 88.24 |
9 | 2084.00 | 117.81 | 2431.25 | 8 | 7.5 | 4.1 | 87.27 | 89.50 |
10 | 2060.31 | 115.56 | 2275.32 | 8 | 7.0 | 2.6 | 86.66 | 88.95 |
11 | 1948.35 | 124.50 | 2522.13 | 8 | 7.0 | 3.0 | 87.28 | 89.55 |
12 | 1927.26 | 128.86 | 2491.21 | 8 | 7.3 | 3.8 | 82.30 | 89.30 |
13 | 1979.14 | 127.64 | 2417.22 | 8 | 7.5 | 3.4 | 86.18 | 88.55 |
14 | 1947.51 | 128.80 | 2568.21 | 8 | 8.0 | 3.5 | 86.85 | 91.35 |
15 | 2075.26 | 120.25 | 2212.54 | 8 | 8.0 | 3.2 | 89.46 | 90.25 |
16 | 2034.28 | 125.28 | 2148.25 | 8 | 7.6 | 3.0 | 90.93 | 91.22 |
17 | 2024.15 | 125.36 | 2044.21 | 8 | 7.6 | 3.7 | 84.68 | 87.57 |
18 | 1974.26 | 115.23 | 1850.24 | 8 | 7.2 | 3.6 | 84.62 | 88.37 |
19 | 1922.92 | 126.50 | 1895.21 | 8 | 7.5 | 4.2 | 87.26 | 89.25 |
20 | 2045.28 | 122.93 | 2412.52 | 8 | 7.0 | 4.5 | 87.25 | 86.80 |
21 | 2088.15 | 112.20 | 2352.01 | 8 | 7.0 | 4.4 | 85.64 | 88.25 |
22 | 2077.48 | 129.64 | 2132.55 | 8 | 7.8 | 4.0 | 89.53 | 91.60 |
23 | 2057.84 | 122.36 | 2045.28 | 8 | 7.6 | 4.1 | 90.38 | 82.25 |
24 | 1948.58 | 118.91 | 2085.33 | 8 | 7.0 | 3.4 | 84.37 | 86.26 |
25 | 1962.45 | 118.55 | 2165.21 | 8 | 7.0 | 4.5 | 85.51 | 88.78 |
26 | 2066.21 | 119.58 | 2158.32 | 8 | 7.6 | 3.8 | 87.50 | 89.28 |
27 | 1925.24 | 124.68 | 2024.32 | 8 | 7.4 | 4.0 | 86.57 | 88.24 |
28 | 1915.24 | 127.58 | 2175.33 | 8 | 7.0 | 2.8 | 85.33 | 88.50 |
29 | 2042.54 | 129.85 | 1950.23 | 8 | 7.8 | 3.0 | 89.28 | 90.26 |
Data set for ANN Model “C” under (
Serial number | Initial COD (mg/L) | Initial |
Initial MLVSS |
Time |
pH | Initial DO |
% COD |
% |
---|---|---|---|---|---|---|---|---|
1 | 1914.23 | 115.65 | 2155.46 | 9 | 7.8 | 3.8 | 94.45 | 88.52 |
2 | 2057.15 | 118.54 | 1844.22 | 9 | 8.0 | 4.4 | 90.11 | 89.82 |
3 | 1982.65 | 124.36 | 1863.51 | 9 | 8.0 | 4.2 | 93.60 | 90.41 |
4 | 1963.22 | 123.43 | 2282.33 | 9 | 7.0 | 3.9 | 95.37 | 90.64 |
5 | 2050.22 | 121.55 | 2072.24 | 9 | 7.0 | 4.1 | 90.33 | 87.73 |
6 | 1978.08 | 122.21 | 2212.52 | 9 | 7.2 | 4.5 | 93.21 | 91.21 |
7 | 1965.47 | 120.33 | 2475.44 | 9 | 7.6 | 3.8 | 90.19 | 88.93 |
8 | 2044.12 | 124.26 | 1840.22 | 9 | 7.2 | 3.0 | 92.08 | 88.35 |
9 | 2084.00 | 127.88 | 2331.25 | 9 | 7.7 | 3.1 | 89.25 | 87.44 |
10 | 2070.31 | 117.54 | 2175.32 | 9 | 7.1 | 3.6 | 89.65 | 86.45 |
11 | 1978.35 | 120.56 | 2422.13 | 9 | 7.4 | 4.0 | 95.22 | 92.54 |
12 | 1967.26 | 124.87 | 2591.21 | 9 | 7.5 | 2.8 | 95.39 | 91.33 |
13 | 1949.14 | 124.67 | 2517.22 | 9 | 7.8 | 4.4 | 94.11 | 92.48 |
14 | 1977.51 | 125.82 | 2068.21 | 9 | 8.0 | 4.5 | 93.05 | 90.61 |
15 | 2075.26 | 119.26 | 2112.54 | 9 | 8.0 | 4.2 | 92.42 | 89.62 |
16 | 2034.28 | 124.29 | 2248.25 | 9 | 7.6 | 4.0 | 94.94 | 89.77 |
17 | 2044.15 | 123.34 | 2244.21 | 9 | 7.7 | 2.7 | 94.18 | 87.65 |
18 | 2084.26 | 125.26 | 1950.24 | 9 | 7.6 | 3.6 | 94.62 | 90.43 |
19 | 2032.92 | 128.52 | 1895.21 | 9 | 7.6 | 3.2 | 95.21 | 91.88 |
20 | 2045.28 | 129.92 | 2112.52 | 9 | 7.2 | 3.5 | 95.22 | 92.95 |
21 | 2078.15 | 116.26 | 2052.01 | 9 | 7.2 | 3.4 | 93.64 | 88.22 |
22 | 2057.48 | 126.66 | 2332.55 | 9 | 7.9 | 3.0 | 89.52 | 86.65 |
23 | 2027.84 | 123.34 | 2545.28 | 9 | 7.5 | 3.1 | 92.39 | 89.24 |
24 | 1948.58 | 119.94 | 1985.33 | 9 | 7.5 | 4.4 | 95.35 | 91.54 |
25 | 2065.25 | 128.92 | 2365.21 | 9 | 7.0 | 4.0 | 91.36 | 90.28 |
26 | 2066.21 | 118.53 | 1958.32 | 9 | 7.3 | 4.2 | 90.55 | 85.32 |
27 | 1945.24 | 124.61 | 1824.32 | 9 | 7.0 | 3.0 | 92.56 | 89.45 |
28 | 1955.24 | 126.56 | 2275.33 | 9 | 8.0 | 3.8 | 88.34 | 85.64 |
29 | 2022.54 | 122.80 | 2550.23 | 9 | 8.0 | 4.0 | 91.22 | 89.65 |
During the training process, small weights were assigned to the connection between neurons in a random way. The weights were modified until the error between the predicted and experimental values of COD and ammonia nitrogen removal efficiency in SBR is minimized. The feedforward backpropagation (BP) algorithm with Levenberg-Marquardt (LM) training was applied for development of all three ANN models. The BP is an approximate steepest descent algorithm with MSE used as performance function. In the neural network development, different number of hidden layers, number of neurons in each layer, and type of transfer function for each neuron were analyzed with a learning rate of 1.0 and training goal of 10−5. Then, the trained networks were tested using the testing data sets and MSE method by modifying the network weights. It was found that network with one hidden layer of neurons was successful. The tansig transfer function was used in the hidden layer and linear transfer function in the output layer. The training of the network was carried out with different number of neurons in the hidden layer with training goal of 10−5. It is observed from Figure
Relation between MSE and number of neurons in hidden layer.
A regression analysis of the network response between the output and the corresponding target was also performed with 18 hidden neurons layer. Figures
ANN Model “A” training, validation, and test squared error for the Levenberg-Marquardt algorithm for prediction of COD and
ANN Model “B” training, validation, and test squared error for the Levenberg-Marquardt algorithm for prediction of COD and
ANN Model “C” training, validation, and test squared error for the Levenberg-Marquardt algorithm for prediction of COD and
The scattered diagrams of linear regression analysis for three model were plotted in Figures
Linear regression analysis of ANN Model “A” for COD removal efficiency during
Linear regression analysis of ANN Model “A” for
Linear regression analysis of ANN Model “B” for COD removal efficiency during
Linear regression analysis of ANN Model “B” for
Linear regression analysis of ANN Model “C” for COD removal efficiency during
Linear regression analysis of ANN Model “C” for
The coefficient of determination is also presented as
The linear regression between the network outputs and the corresponding targets showed that the neural network outputs (forecasted data) were obviously agreed with the experimental values. The correlation between ANN testing outputs and the experimental values of COD and
Comparison of ANN Model “A” output and experimental values for test data set.
Serial number | COD removal efficiency (%) |
|
Error (%) | |||
---|---|---|---|---|---|---|
Experimental values | ANN predicted values | Experimental values | ANN predicted values | COD |
|
|
1 | 86.45 | 87.32 | 86.92 | 85.54 | −0.87 | 1.38 |
2 | 89.33 | 90.52 | 85.83 | 87.22 | −1.19 | −1.39 |
3 | 88.65 | 86.15 | 82.97 | 84.19 | 2.50 | −1.22 |
4 | 90.42 | 88.95 | 84.22 | 85.12 | 1.47 | −0.90 |
5 | 85.22 | 84.55 | 82.85 | 82.12 | 0.67 | 0.73 |
6 | 84.29 | 87.15 | 85.52 | 84.28 | −2.86 | 1.24 |
7 | 90.22 | 89.51 | 88.25 | 86.45 | 0.71 | 1.80 |
Comparison of ANN Model “B” output and experimental values for test data set.
Serial number | COD removal efficiency (%) |
|
Error (%) | |||
---|---|---|---|---|---|---|
Experimental values | ANN predicted values | Experimental values | ANN predicted values | COD |
|
|
1 | 82.35 | 83.18 | 86.45 | 85.15 | −0.83 | 1.30 |
2 | 85.58 | 84.72 | 88.24 | 87.27 | 0.86 | 0.97 |
3 | 82.30 | 84.15 | 89.30 | 90.56 | −1.85 | −1.26 |
4 | 90.93 | 91.25 | 91.22 | 88.95 | −0.32 | 2.27 |
5 | 87.25 | 86.12 | 86.80 | 88.16 | 1.13 | −1.36 |
6 | 84.37 | 82.78 | 86.26 | 85.75 | 1.59 | 0.51 |
7 | 85.33 | 84.62 | 88.50 | 90.78 | 0.71 | −2.28 |
Comparison of ANN Model “C” output and experimental values for test data set.
Serial number | COD removal efficiency (%) |
|
Error (%) | |||
---|---|---|---|---|---|---|
Experimental values | ANN predicted values | Experimental values | ANN predicted values | COD |
|
|
1 | 93.60 | 95.24 | 90.41 | 88.35 | −1.64 | 2.06 |
2 | 90.19 | 92.85 | 88.93 | 90.18 | −2.66 | −1.25 |
3 | 95.22 | 93.28 | 92.54 | 90.76 | 1.94 | 1.78 |
4 | 94.18 | 95.15 | 87.65 | 88.25 | −0.97 | −0.60 |
5 | 95.21 | 91.88 | 91.88 | 92.85 | 3.33 | −0.97 |
6 | 92.39 | 90.74 | 89.24 | 90.12 | 1.65 | −0.88 |
7 | 92.56 | 91.25 | 89.45 | 87.48 | 1.31 | 1.97 |
Comparison of performance statistics of three ANN models.
Model number | Pollutant removal efficiency | ME | MSE | RMSE | Coefficient of determination | Slope |
|
---|---|---|---|---|---|---|---|
Model “A” | COD removal |
0.06 |
2.81 |
1.67 |
0.960 |
0.918 |
0.0537 |
| |||||||
Model “B” | COD removal |
0.18 |
1.32 |
1.14 |
0.947 |
0.928 |
0.0494 |
| |||||||
Model “C” | COD removal |
0.42 |
4.28 |
2.06 |
0.955 |
0.932 |
0.0495 |
Stimulation results of Model “A” for test data sets.
Stimulation results of Model “B” for test data sets.
Stimulation results of Model “C” for test data sets.
Based on performing all relevant experiment and analysis of results, it is concluded that SBR is a reasonable alternative option for simultaneous removal of high COD and
The three neural network Models “A,” “B,” and “C” under three different combinations of aerobic-anoxic sequence, namely,
The authors disclosed the fact that this paper has been prepared for academic purpose for research persuasion and exchange of knowledge. There is no bearing on any financial relation with any commercial identities and also avoiding any conflict of interests.
This study was supported by the research funds of Jadavpur University, Jadavpur, Kolkata-32.