A Research On The Medium-&Long Term Load Prediction Methods And Electricity Purchasing Mode On Electricity Market In Quanzhou

Aiming at the issues of the medium-&long term load prediction methods and electricity Purchasing mode on Electricity Market in Quanzhou, the thesis analyzes in detail the economic development, electricity consumption, load features and electricity purchase of Quanzhou, builds up corresponding models for the medium-&long term load prediction methods and electricity Purchasing mode . Since Quanzhou is located in the golden triangle of South Fujian, its economy develops fast, leading to a rapid load growth. Regarding the medium-&long term load prediction, the thesis first comes up with an improved grey prediction model, also called improved GM(1,1) load prediction model with white exponential law coincidence, betters the precision of grey prediction, widens its application through analyzing the restriction common grey prediction model; Secondly, the thesis constructs NPLS load prediction model based on analyzing the mechanisms of linear PLS model. Though examples, two approaches adopted are proven to be more adaptable to the load prediction of Quanzhou and proven to reach satisfying results. The thesis combines these two approaches to have got a better application. Since Quanzhou is still on the elementary stage of electricity market , Regarding electricity Purchasing mode,the thesis through analyzing the actual situation of electricity sources in Quanzhou, builds up annual purchase model, per-month purchas model and pre-day purchase model. In the annual contracted purchase model ,two section biding mechanism is adopted to construct annual construct-market purchase model, mainly to have resolved the problem of the capacity charges and the volume charges. the electricity volume will be decided according to the result of load prediction in a annually decreasing way, to resolve historical problems so that the machine sets could be on the same scratch in the pre-day market biding, and then the contracted electricity volume could be gradually reduced leading to the transformation to a complete price biding. In the pre-month market biding, the restriction of annual contracted volume is considered to have made the purchase cost lowest. Finally, the pre-month volume is reasonably allocated to the pre-day market, together with the annual contracted volume. In the pre-day market purchase model, the machine sets launches biding according to the principle that the lower-priced electricity should be allocated first, the settlement is conducted by sectional uniform clearing price with consideration that purchase cost can be made lowest on the occasion where restrictive conditions should be satisfied. The thesis emphatically considers blocking biding mechanism of economic allocation for the contracted volume in the pre-day market, resolves the problem that previous block biding didn't consider day-contract volume allocation, proves the adopted method in the thesis is much more suitable to the purchase decision in Quanzhou. In the end, the thesis provides example analysis.