Abstract

There are some theories on how the World will end (e.g. super volcano eruption, massive star explosion, death of the Sun, asteroid impact, pandemic, nuclear war, climate change). Some of them can be prevented, because the cause is human by itself. For instance, spread of deadly diseases can be prevented by some quarantine zones and periods, nuclear wars by disarmament of weapon of mass destruction (zero weapons) and climate change by new life styles and acts (zero emissions: carbon dioxide CO2, methane CH4, nitrous oxide N2O, fluorinated gases). Electricity generation plays a key role in zero emissions life styles and acts. A Global Grid can be designed, invested and operated by 100% renewable energy power plants on the World. Design and operation of this grid needs some very detailed electricity demand information. One of this information is the long term electricity demand prediction (PWh: Petawatt hours). This paper investigates an experimental Mamdani's type fuzzy inference system for the Global Grid electricity demand forecasting in this respect. Two, three, four, five, six, seven, eight, and nine (2 to 9) triangular membership functions and respective Mamdani's rules are modeled in a systematic manner and tested and finally presented in this study. Maximum absolute percentage errors (MAP) are respectively calculated as 0,66, 0,65, 0,52, 0,42, 0,35, 0,32, 0,33, and 0,32. Mean absolute percentage errors (MAPE) are 0,49, 0,53, 0,37, 0,30, 0,28, 0,27, 0,26, and 0,26. This research paper will hopefully be a good start for a worldwide research, development, demonstration,