{"id":{"repo_id":"texas-state","oai_identifier":"oai:digital.library.txst.edu:10877/24764"},"canonical_url":"https://search.dev.ndltd.org/etd/texas-state/oai:digital.library.txst.edu:10877/24764","repository":{"repo_id":"texas-state","name":"Texas State University","base_url":"https://digital.library.txst.edu/server/oai/request"},"display":{"title":"Deploying EV Charging Infrastructure using Optimization with Real-Life Data","abstract":"As electric vehicle (EV) adoption accelerates, the strategic deployment of charging infrastructure in parking facilities presents both an opportunity and a planning challenge for operators. This research develops and validates optimization models for Level 2 EV charging station deployment and pricing in parking garages, utilizing a novel API-available parking occupancy dataset from Santa Monica spanning 2018-2022. Three optimization approaches— GridSearch, Gurobi mixed-integer programming with decision trees, and OptiCL machine learning—are employed to jointly determine optimal charging prices and infrastructure capacity across three parking garages representing diverse demand environments: recreational beach traffic, commercial office, and mixed urban use. The analysis makes four primary contributions. First, statistical analysis of pre-pandemic (2018-2019) and post-pandemic (2021-2022) parking behavior reveals significant demand pattern shifts attributable to COVID-19, with regression models confirming that pandemic severity metrics significantly predicted parking availability changes. Second, temporal optimization methods incorporating hourly and day-of-week demand patterns substantially outperform traditional single-value approaches: baseline optimizations using mean or mode demand recommended 50-70% fewer chargers and achieved 20-150% lower simulated profits compared to temporal methods. Third, all three optimization approaches converge on nearly identical solutions—within 0.04 on optimal price and identical charger counts—providing confidence in solution robustness while revealing computational trade-offs, with Gurobi solving 100-200x faster than alternatives. Fourth, sensitivity analysis across time intervals (5 to 120 minutes) demonstrates that coarser temporal resolutions provide sufficient accuracy for practical planning while substantially reducing computational burden. Historical simulation validation against actual parking data revealed systematic but modest prediction errors of -0.7% to -8.7% depending on garage characteristics, with higher-utilization facilities exhibiting greater predictability. Optimal configurations varied substantially across locations: the high-demand Civic Center garage supported 14 chargers at 3.88-3.90 with simulated daily profits of 428, while the lower-demand Beach House garage optimized at 7 chargers and 1.86 with 79 daily profit. These findings demonstrate that data-driven optimization using readily available parking behavior data can identify profitable EV charging configurations tailored to specific location characteristics, providing practical guidance for parking operators considering infrastructure investments.","abstract_html":"As electric vehicle (EV) adoption accelerates, the strategic deployment of charging infrastructure in parking facilities presents both an opportunity and a planning challenge for operators. This research develops and validates optimization models for Level 2 EV charging station deployment and pricing in parking garages, utilizing a novel API-available parking occupancy dataset from Santa Monica spanning 2018-2022. Three optimization approaches— GridSearch, Gurobi mixed-integer programming with decision trees, and OptiCL machine learning—are employed to jointly determine optimal charging prices and infrastructure capacity across three parking garages representing diverse demand environments: recreational beach traffic, commercial office, and mixed urban use. The analysis makes four primary contributions. First, statistical analysis of pre-pandemic (2018-2019) and post-pandemic (2021-2022) parking behavior reveals significant demand pattern shifts attributable to COVID-19, with regression models confirming that pandemic severity metrics significantly predicted parking availability changes. Second, temporal optimization methods incorporating hourly and day-of-week demand patterns substantially outperform traditional single-value approaches: baseline optimizations using mean or mode demand recommended 50-70% fewer chargers and achieved 20-150% lower simulated profits compared to temporal methods. Third, all three optimization approaches converge on nearly identical solutions—within 0.04 on optimal price and identical charger counts—providing confidence in solution robustness while revealing computational trade-offs, with Gurobi solving 100-200x faster than alternatives. Fourth, sensitivity analysis across time intervals (5 to 120 minutes) demonstrates that coarser temporal resolutions provide sufficient accuracy for practical planning while substantially reducing computational burden. Historical simulation validation against actual parking data revealed systematic but modest prediction errors of -0.7% to -8.7% depending on garage characteristics, with higher-utilization facilities exhibiting greater predictability. Optimal configurations varied substantially across locations: the high-demand Civic Center garage supported 14 chargers at 3.88-3.90 with simulated daily profits of 428, while the lower-demand Beach House garage optimized at 7 chargers and 1.86 with 79 daily profit. These findings demonstrate that data-driven optimization using readily available parking behavior data can identify profitable EV charging configurations tailored to specific location characteristics, providing practical guidance for parking operators considering infrastructure investments.","abstract_has_math":false,"creators":["Mallett, Elizabeth"],"institution":"Texas State University","degree_name":"Master of Science","degree_level":"Masters","degree_discipline":"Data Analytics and Information Systems","degree_department":null,"school":null,"contributors":[],"advisors":["Zhu, Emily"],"committee_chairs":[],"committee_members":["Akire, Linda","Mendez, Francis","Anderson, Sid"],"year":2026,"date_issued":"2026-05","date_published":"2026-05","updated_at":"2026-07-27T21:22:45Z","subjects":["EV charging","electric vehicle","charging","optimization","optimization with real data","parking demands","EV charging demands","charging demands","parking behaviors"],"languages":["en"],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/10877/24764","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Zhu, Emily"]},{"key":"dc:contributor.committeemember","label":"Committee Member","values":["Akire, Linda","Mendez, Francis","Anderson, Sid"]},{"key":"dc:creator","label":"Author","values":["Mallett, Elizabeth"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-05-11T15:25:18Z"]},{"key":"dc:date.issued","label":"Date","values":["2026-05"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Data Analytics and Information Systems"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Masters"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["Texas State University"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["EV charging","electric vehicle","charging","optimization","optimization with real data","parking demands","EV charging demands","charging demands","parking behaviors"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/10877/24764"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["As electric vehicle (EV) adoption accelerates, the strategic deployment of charging infrastructure in parking facilities presents both an opportunity and a planning challenge for operators. This research develops and validates optimization models for Level 2 EV charging station deployment and pricing in parking garages, utilizing a novel API-available parking occupancy dataset from Santa Monica spanning 2018-2022. Three optimization approaches— GridSearch, Gurobi mixed-integer programming with decision trees, and OptiCL machine learning—are employed to jointly determine optimal charging prices and infrastructure capacity across three parking garages representing diverse demand environments: recreational beach traffic, commercial office, and mixed urban use. The analysis makes four primary contributions. First, statistical analysis of pre-pandemic (2018-2019) and post-pandemic (2021-2022) parking behavior reveals significant demand pattern shifts attributable to COVID-19, with regression models confirming that pandemic severity metrics significantly predicted parking availability changes. Second, temporal optimization methods incorporating hourly and day-of-week demand patterns substantially outperform traditional single-value approaches: baseline optimizations using mean or mode demand recommended 50-70% fewer chargers and achieved 20-150% lower simulated profits compared to temporal methods. Third, all three optimization approaches converge on nearly identical solutions—within 0.04 on optimal price and identical charger counts—providing confidence in solution robustness while revealing computational trade-offs, with Gurobi solving 100-200x faster than alternatives. Fourth, sensitivity analysis across time intervals (5 to 120 minutes) demonstrates that coarser temporal resolutions provide sufficient accuracy for practical planning while substantially reducing computational burden. Historical simulation validation against actual parking data revealed systematic but modest prediction errors of -0.7% to -8.7% depending on garage characteristics, with higher-utilization facilities exhibiting greater predictability. Optimal configurations varied substantially across locations: the high-demand Civic Center garage supported 14 chargers at 3.88-3.90 with simulated daily profits of 428, while the lower-demand Beach House garage optimized at 7 chargers and 1.86 with 79 daily profit. These findings demonstrate that data-driven optimization using readily available parking behavior data can identify profitable EV charging configurations tailored to specific location characteristics, providing practical guidance for parking operators considering infrastructure investments."]},{"key":"dc:format","label":"Dc Format","values":["Text"]},{"key":"dc:format.medium","label":"Dc Format Medium","values":["1 file (.pdf)"]},{"key":"dc:title","label":"Title","values":["Deploying EV Charging Infrastructure using Optimization with Real-Life Data"]}]}],"canonical_facts":{"dc:contributor.advisor":["Zhu, Emily"],"dc:contributor.committeemember":["Akire, Linda","Mendez, Francis","Anderson, Sid"],"dc:creator":["Mallett, Elizabeth"],"dc:date.accessioned":["2026-05-11T15:25:18Z"],"dc:date.issued":["2026-05"],"dc:description.abstract":["As electric vehicle (EV) adoption accelerates, the strategic deployment of charging infrastructure in parking facilities presents both an opportunity and a planning challenge for operators. This research develops and validates optimization models for Level 2 EV charging station deployment and pricing in parking garages, utilizing a novel API-available parking occupancy dataset from Santa Monica spanning 2018-2022. Three optimization approaches— GridSearch, Gurobi mixed-integer programming with decision trees, and OptiCL machine learning—are employed to jointly determine optimal charging prices and infrastructure capacity across three parking garages representing diverse demand environments: recreational beach traffic, commercial office, and mixed urban use. The analysis makes four primary contributions. First, statistical analysis of pre-pandemic (2018-2019) and post-pandemic (2021-2022) parking behavior reveals significant demand pattern shifts attributable to COVID-19, with regression models confirming that pandemic severity metrics significantly predicted parking availability changes. Second, temporal optimization methods incorporating hourly and day-of-week demand patterns substantially outperform traditional single-value approaches: baseline optimizations using mean or mode demand recommended 50-70% fewer chargers and achieved 20-150% lower simulated profits compared to temporal methods. Third, all three optimization approaches converge on nearly identical solutions—within 0.04 on optimal price and identical charger counts—providing confidence in solution robustness while revealing computational trade-offs, with Gurobi solving 100-200x faster than alternatives. Fourth, sensitivity analysis across time intervals (5 to 120 minutes) demonstrates that coarser temporal resolutions provide sufficient accuracy for practical planning while substantially reducing computational burden. Historical simulation validation against actual parking data revealed systematic but modest prediction errors of -0.7% to -8.7% depending on garage characteristics, with higher-utilization facilities exhibiting greater predictability. Optimal configurations varied substantially across locations: the high-demand Civic Center garage supported 14 chargers at 3.88-3.90 with simulated daily profits of 428, while the lower-demand Beach House garage optimized at 7 chargers and 1.86 with 79 daily profit. These findings demonstrate that data-driven optimization using readily available parking behavior data can identify profitable EV charging configurations tailored to specific location characteristics, providing practical guidance for parking operators considering infrastructure investments."],"dc:format":["Text"],"dc:format.medium":["1 file (.pdf)"],"dc:identifier.uri":["https://hdl.handle.net/10877/24764"],"dc:language.iso":["en"],"dc:subject":["EV charging","electric vehicle","charging","optimization","optimization with real data","parking demands","EV charging demands","charging demands","parking behaviors"],"dc:title":["Deploying EV Charging Infrastructure using Optimization with Real-Life Data"],"dc:type":["Thesis"],"thesis:degree_discipline":["Data Analytics and Information Systems"],"thesis:degree_level":["Masters"],"thesis:degree_name":["Master of Science"],"thesis:institution_name":["Texas State University"]},"updated_at":"2026-07-27T21:22:45Z"}