BIOINSPIRED ALGORITHM FOR ESTIMATING DRYING PARAMETERS FOR PLANTAIN.pptx

BIOINSPIRED ALGORITHM

1. BIOINSPIRED ALGORITHM FOR ESTIMATING DRYING PARAMETERS FOR PLANTAIN

2. OUTLINE BACKGROUND OF STUDY STATEMENT OF PROBLEM OBJECTIVE OF THE STUDY SIGNIFICANCE OF THE STUDY SCOPE AND LIMITATION LITERATURE REVIEW MATERIALS AND METHODS RESULTS AND DISCUSSIONS CONCLUSION AND RECOMMENDATION

3. BACKGROUND OF STUDY As a unit operation, drying of solid materials is one of the most common and important in the chemical process industries (CPI). The effectiveness of drying processes can have a large impact on product quality and process efficiency. Dried foods can be stored for long periods without deterioration occurring. Microorganisms which cause food spoilage and decay are unable to grow and multiply in the absence of sufficient water. Drying is essential to reduce post harvest losses in the cultivation ad marketing of plantain.

4. STATEMENT OF THE PROBLEM Post harvest losses Determining appropriate values to store plantain for longer periods Storage and preservation problem

5. The specific objectives are to: 1. To determine the drying parameters (drying rate, moisture content) of plantain. 2. Prediction of drying rate and moisture content using artificial neural network. OBJECTIVE OF THE STUDY

6. The aim of this study was to solve the storage and preservation problem of plantain, by reducing losses. The research work would serve as a base for further research work. The developed model helped to determine the moisture ratio of the plantain sample . SIGNIFICANCE OF THE STUDY

7. SCOPE AND LIMITATION The scope of the study is listed as follows: 1. Identification and collection of the raw material 2. Preparation of the raw material sample 3. Drying of the sample 4. Prediction of the drying parameters using Artificial Neural Network (ANN).  Limitation of the study: The work was limited to the drying of unripe plantain ( Musa paradisiaca ) found in Nigeria and the prediction of drying parameters using ANN model.

8. LITERATURE REVIEW Type of product Method used Outcome Input parameters Output parameters green pea microwave assisted fluidized bed dryer network with the logsig (Log sigmoid) transfer function and train rp (Resilient back propagation; Rprop ) back propagation algorithm made the most accurate predictions for the green pea drying system Microwave power, drying air temperature, moisture content drying time ginkgo biloba seeds microwave drying ANN methodology could precisely predict experimental data with high correlation coefficient microwave power, drying time moisture content banana drying process multi-layer perceptron network is better than mathematical equations to predict the experimental data because of its universality air temperature, air velocity, drying time and slices thickness moisture content

9. MATERIALS AND METHODS Preparation of Samples Drying experiments Data acquisition and preparation Artificial Neural Network (ANN) Training Design of Network Architecture Artificial Neural Network (ANN) Validation and Testing

10. RESULTS AND DISCUSSIONS Figure 4.1: Network Training Performance

11. Figure 4.2 Regression analysis plot

12. W 1 ,I,th W 1 ,I,T W 1 ,I,t Σ f Thickness Temperature Time Drying rate Moist ure ratio Figure 4.3 ANN result topology

13. Figure 4.4 Network training algorithm

14. Figure 4.5: Predicted network architecture

15. Table 4.1: Network Weights and Biases

16. Input parameters Time, Temperature, slice thickness Output parameter Moisture Ratio (MR) Partition ratio 70-30-30 Network topology 3-4-2 Training algorithm Levenberg Marquardt Network type Feed forward back propagation Transfer function logsig R 0.97858 MSE 0.0046605 Table 4.2: Summary of results

17. CONCLUSION AND RECOMMENDATION The predictions were very close to the experimental data, with the MSE value 0.0046605 at the 39th epoch and R value of 0.97858. The output of the model provided good matches with original experimental data and the work shows the reliability and applicability of the ANN in prediction of drying parameters. There are several other unexplored models and further research work is recommended to be carried out determine if any of them could possibly describe the drying of plantain better.