FCP dataset for forecasting temperature, PV, price, and load

Singapore aims to transform into a green and sustainable city by 2030. One of the key actions is to phase out Internal Combustion Engine (ICE) vehicles and significantly expand electric vehicle (EV) adoption. An EV is powered by electricity generated from natural gas and renewables, so the average carbon emission is only half of an ICE vehicle powered by petrol and diesel. By 2030, Singapore will cease the new registration for ICE cars, and eventually, all vehicles will run on clean energy by 2040. With the massive expansion of EVs in the foreseeable future, EV charger installation shall also match the trend and there will be at least 60,000 EV chargers deployed by 2030, roughly five EVs per charger. The action is ambitious indeed, and the EV and charger network system is expected to be enormous soon.<br>The rapid expansion of the system comes with the requirements of advanced management and operation. However, EV chargers nowadays cannot well satisfy the requirements. The key issue is that EV chargers are not smart. Largely due to the cost consideration, an EV charger is more of an electricity transforming and delivery unit instead of a computation-driven intelligent module, and computational resource is often missing or minimal in existing chargers. Besides computing resources, EV chargers also lack sufficient capability for connectivity, where data transmission is mostly cable-based for wired transmission and only relatively advanced chargers support wireless connectivity like 4G and Wi-Fi. Lack of intelligence with data scarcity might be acceptable for early-stage small-scale deployment. But for a large-scale system, potential consequences can be poor management, inferior scheduling, economic loss, weekend reliability, and so on.<br>In this project, we propose to empower the EV chargers with 5G capabilities for connectivity and computing and bring smartness and intelligence into them. 5G is fast, so the high-resolution EV charger data can be accessed in real-time with minimal delay. 5G supports high concurrency, so a large number of EV chargers can utilize the connectivity without being forced to be sequential to avoid conflict and long delay. 5G has great bandwidth, so abundant information from EV chargers and the associated facilities like battery energy storage systems (BESS) and solar panels can be transmitted. 5G is also ultra-reliable with low latency which makes 5G suitable for mission critical functionalities and time-sensitive control. Overall, 5G connectivity addresses the key challenge of data scarcity in current chargers and facilitates data-driven system monitoring and intelligent management. Besides providing connectivity, 5G is also featured with edge computing capability with edge servers integrated into 5G networks. So, the data can be processed and analyzed in edge servers, where the computing resource enables insights and knowledge extraction from the data to realize intelligent EV charging management. To achieve the overall goal of the 5G-powered intelligent EV charging system, we have the following key objectives for our research.<br>• To design and develop 5G-based data processing and analytics systems and interfaces for data acquisition, transmission, storage, management, and analytics.• To design and develop data-driven algorithms for accurate and reliable charging supplydemand forecasting and cost-optimal scheduling with large-volume and high-resolution data.• To implement a prototype and demonstrate the system effectiveness utilizing the facilities from SIT’s Future Communications Translation Lab (FCTLab) and our EV sector industry partner. Upon successful demonstration, our industry partner plans to commercialize the solutions and deploy them in the company’s EV charging system for widespread adoption.<br>Our research tackles the urgent challenges of lacking connectivity, hence data-driven intelligence in the current industry of EV charging management. We leverage the 5G capabilities of connectivity and computation to promote data availability and analytics. We believe our research is promising with strong support from both academia and industry. The research has a significant impact on upgrading the EV or mobility industry with great potential for economic and sustainability.<br><br>1. Source of Weather Dataset: https://www.visualcrossing.com/2. Source of PV Dataset: https://purl.stanford.edu/fb002mq94073. Source of Price Dataset: https://www.nems.emcsg.com/nems-prices4. Source of EV Charging Demand Dataset:https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/NFPQLW5. Source of EV Charging Demand Dataset: https://data.cityofpaloalto.org/dataviews/257812/electric-vehiclecharging-station-usage-july-2011-dec-2020/<br><br>