Power Consumption Prediction Project Presentation

This presentation dives into the methodologies and tools used for predicting power consumption. Tailored for students, it covers the importance of power consumption forecasting, various prediction techniques, data requirements, and practical applications in energy management and sustainability.
for more visit: https://bostoninstituteofanalytics.org/data-science-and-artificial-intelligence/

1. POWER CONSUMPTION PREDICTION -Chaitanya Tajne

2.  To create a machine learning model capable of accurately predicting power consumption in Zone 3.  Deliver the insights from the data Significant parameters to be monitor AGENDA

3. Data Pre-Processing Feature engineering Eda Model Building Model Deployment(Web App) POWER CONSUMPTION PREDICTION

4. DATA PRE-PROCESSING Check Data Types Dropped Duplicates Checked Null Values Outliers Detection Outlier Removal By Using IQR Method

5. FEATURE ENGINEERING Extracted features from date time Extracted features : peak hour, business hour, business day, weekend, week-day, time, hour etc.

6. EDA

7. There is a strong positive correlation between average monthly temperature and average monthly power consumption in Zone 3. The highest average monthly power consumption occurs in July, corresponding to the highest average monthly temperature. The lowest average monthly power consumption occurs in December, corresponding to the lowest average monthly temperature There is a trend of increasing power consumption from January to July, followed by a decline from July to December This trend suggests that people in Zone 3 tend to use more energy during the warmer months, possibly for air conditioning or other cooling purpose The data also shows a significant drop in power consumption from August to September, which could be attributed to factors like seasonal changes in weather patterns, reduced daylight hours, or changes in consumer behave Overall, the data provides valuable insights into the relationship between temperature and power consumption, highlighting the potential impact of climate change on energy demand.

8. The graph shows an inverse relationship between average monthly humidity and average monthly power consumption in Zone 3. As humidity increases, power consumption decreases.- This trend is particularly evident between Month 7 and Month 10, where humidity increases significantly, and power consumption drops. The highest average monthly humidity was observed in Month 4 with 74% and the lowest was observed in Month 12 with 13.3%. The highest average monthly power consumption was observed in Month 8 with 27,000 kWh and the lowest was observed in Month 12 with 11,000 kWh. The months with the highest humidity levels (May-September) generally coincide with the months with the lowest power consumption.

9. The graph displays the relationship between wind speed and average power consumption of zone 3. The average power consumption generally decreases as wind speed increases. However, there are some outliers, where the power consumption is high even with high wind speed. It suggests that there might be other factors affecting power consumption besides wind speed. Further investigation is required to understand the relationship between wind speed and power consumption.

10. The scatter plot shows the relationship between General Diffuse Flows and Average Power Consumption Zone 3. As General Diffuse Flows increase, Average Power Consumption Zone 3 tends to decrease.- The relationship between the two variables appears to be non-linear, with a decreasing trend.- There is a large amount of data points clustered together in the lower right corner of the plot, suggesting that a large portion of the data has high General Diffuse Flows and low Average Power Consumption Zone 3. This plot could be used to understand the relationship between General Diffuse Flows and Average Power Consumption Zone 3 and to identify potential areas for optimization.

11. The scatter plot shows the relationship between diffuse flows and average power consumption of zone 3. The plot suggests that there is a weak positive correlation between the two variables. However, the relationship is not very strong, meaning that changes in diffuse flows do not necessarily lead to significant changes in average power consumption. The data points are clustered in a relatively narrow band, indicating that the power consumption for zone 3 generally remains within a certain range, regardless of the diffuse flow values. The plot could be further analyzed by looking at the distribution of data points within the band to identify any potential patterns or anomalies.

12. Power BI Dashboard

13. MODEL BUILDING Model Train Accuracy (%) Test Accuracy (%) Ada Boost 77.706 78.228 ANN 96.249 96.135 Random Forest 90.317 90.263

14. MODEL DEPLOYMENT Link: https://powerprediction-ddjuo6jspeq8bwpme7uexc.streamlit.app/

15. Questions ?

16. Thank You!