Bi-objective shortest path algorithm

When it comes to mode choice problems, there are number of factors that are taken into account while choosing the optimal mode between two points. Typically, the factors include total travel time, cost of travel, inconvenience, resource consumption and so on, and the importance of each factor may vary from person to person. This project will limit the number of objectives to two, and will compare the optimal mode choices for different income groups for the Bay Area Rapid Transit (BART) network in the San Francisco Bay Area. It will also limit the study for ‘peak hours’, since the off-peak mode competition can give an unfair advantage to driving as the train schedule intervals are wide and no congestion delays happen to driving.

This project will focus on the competition between the three modes of choice (Driving alone, Carpooling, taking transit) for the different income groups ranging from minimum wage to top 5%, for the BART train network in the Bay Area. The objective function is two-fold in this case, it is designed to first minimize the total travel time, and second, it is designed to minimize the total travel cost, and the optimal mode based on these two is then selected.

The shortest path algorithm was developed in Python and the BART network and shortest path were determined using networkx package.

Spatial Renewable Energy Power Plant Locator

This project focuses on the electricity generation sector. It takes the emissions data from an energy systems model, and identifies optimal locations of renewable energy power plants, specifically, (a) Wind power, (b) Solar power, (c) Biomass, and (d) Geothermal. Several geographical factors play into each of the renewable energy power plants. This project finds locations to meet the energy demand considering the respective factors of each type of power plant. Finally, once the locations are determined, the investment costs for each energy source is determined.

ArcGIS package is used for this project with Arcpy module for Python scripting. Numpy package is used for the mathematical calculations in the model. The output is a customized ArcToolBox tool that will take the input file from the model and generate new power plant locations geographically and write the new power plant investment cost output to a text file.

 Schematic of the Spatial Power Plant Locator Algorithm

Schematic of the Spatial Power Plant Locator Algorithm

The model is analyzed for the state of California. Two scenarios were analyzed through this tool: (a) business-as-usual scenario and (b) carbon reduction scenario. Under business-as-usual scenario, there are no carbon targets in the state. Under the carbon reduction scenario, the model is trying to meet the state's AB-32 Global Warming policy of 80% reduction of 1990 greenhouse gas emission levels by 2050.