The worldwide eIectricity market reform brings about new development opport'IJnities forelectric power industry, and also puts forward many new research topics for power systemanalysis. Several key problems in electricity markets, such as pricing of transmission andwheeIing transactions, generation bidding and available transfer capability calculation betWeeninterconnected systems, etC., are comprehensively and thoroughly studied and analyzed in thisdissertation. The main contents and contributions are as follows:(l ) ln a spot price based marketplace, the revenues and costs of transmission service are noteasily balanced due to the economies of scale of transmission netWorks. A novel revenuereconciliation method for spot pricing using power flow tracing is proposed. In the light of useof transmission system, revenue reconciliation is realized by use of a surcharge, which does notchange the spot price signal and hence attains maximum socia1 welfare. Directly relatCd to the' system operation condition, the original spot price can be regarded as an economic signal forsystem short term operation; reflecting the location of each node, surcharge of spot price can beconsidered as an economic signal for systCm long term run. The advantage of the proposedmethod lies in that not only the fair and equitable principle for every market participant isensured, but also the long term economic signaI which cannot be obtained from spot pricing isprovided.(2) Based on the whee1ing rate derived from spot pricing, a wheeling transaction chargingmethod is proposed by use of a surcharge, which can refiect the percentage usage of thewheeIing transaction to the transmission system. The method is performed on the basis of theload flow mode inc1uding a1l wheeling transactions, thus it can consider multipIe wheelingtransactions simultaneously. The proposed method is realized based on technical rather thanlegal assessment, hence the equitable principIe of marketplace can be guaranteed.(3) Based on power flow tracing, a modified merit of order auction method considering thenetwork loss is presented. Due to the differences of location, the generation power from eachpower plant should have different network loss level. Having taken netWork loss effect of each+mdsh#%lRAttM1$tt&xpower plare's active power injection into accouat before market dispatching, the modifiedmethod may generate a more equitable marke clearing results than that of the original merit oforder method. Furthermore, due tO the influence of nforork loss, the market clearing price willdiffer from point to point by use of the new auction method, hence it can be used as a pricesignal to motivate siting of power plants in the long run.(4) For the first time, by use of the reinforcement learning scheme, the model and algorithmtO determine the optimal bidding strategy fOr an independent market participant are proposed.As a locally performed model-free method, it can assist a markst participant to obtain its globaloptimal bidding stratCgy. The proposed learning automata model and bidding algorithm neednot to know the uncertain information in the marketplace, such as the competitors' bidingbehaviors, etc., and need not to conduct complicated optimization tasks, therefore it is very easyto implement, and applicable to the dotermination of individual market participant's biddingstrategy in large scale electricity markets.(5) Taking all kinds of security constraints into account, the mathematical model isestabIished tO analyzing available transfer capability (ATC) of interconnected power systems,and a practical continuation method is propose for calculating ATC. Base on thepredicatOr-correctOr and automatic step size contro1 technique, combined with contingencyscreening and selection, the proposed method possesses high computation efficiency.The above mentioned methods are illustfated by the lEEE 30-bus test system, WSCC 9-bustCSt system and the Northwe