Research On Medium And Long-term Security Constraints Unit Commitment

The solution of day-ahead unit commitment problem gradually matured around the world. Currently the scale of power state grid continues expanding, and the high voltage transmission lines has been put forward much more. That means the more rigorous requirements for security and reliability of power grids. So the long-term security constrained unit commitment (SCUC) has become the hotspot at this stage. The purposes of this paper are to promote the optimization technology over different time scales, create a set of elementary system suitable for Chinese national conditions, and build the corresponding software for engineering.Firstly the authors established the power element models based on physical and systemic properties of the power components, including common thermal units, combined cycle gas turbines units, pumped-storage hydro units, storage hydro units and tie lines. Meanwhile they modeled the power plants for the sum of units’generated energy, because in China we were concerned about the energy of each plant, rather than each unit. And consdidering the relevant constraints of system operation, they presented a computationally efficient mixed integer linear formulation for the SCUC problem of multi-type units. Moreover, the modal transfer matrix proposed in combined cycle gas turbine modeling could be applied to any states constraint modeling of any units, so the systematic modeling will be easier.The long-term SCUC focuses on the economic operation during longer period, the more precise results, and the higher computing speed. After analyzing the particularity of long-term SCUC, the authors efficiently built the special relative models for time coupling constraints, power rates and energy rates recoupling constraints, and grid security constraints. In the time coupling constraint analysis, they referred to the cutting plane method, used inequality magnification and shrinking techniques, and adjusted the constraint-variable correlation of valid inequality constraints. And the numerical results showed that it would narrow the initial feasible region, and get the optimal solution rapidly. In recoupling on power rates and schedule energy rates, they adopted the idea of smooth ramping to limit the schedule energy rates variables, which would establish the relationship between the power outputs and schedule energy again. The numerical results convinced that the new formulation could ensure the calculation efficiency and improve the credibility of the solution. In modeling the grid security constraints, they proposed polymorphic checks with the same interval time, which separated the check procedures of the base security and the N-1 security. The base one should be limited all over the periods, while the N-1 would be checked just the former several days, and scrolling another several days to refresh the check results in next optimization. The reduction in the computational burden decreased the computing time required by available commercial software to solve the problem. It is worth mentioning that they also constructed the grid security checks with the different interval time.The proposed model has been successfully tested on dual cases of IEEE 118 case and a Chinese regional grid case. Numerical results have revealed the accurate and efficient performance of the new formulation. IEEE118 case proved that the economic bebefits of multi-type units cooperation was superior to homegeneous thermal units, and then the causes of optimization results were analyzed. In the regional power grid case, all plants appeared pretty much meeting the generated energy schedule, and almost all energy rates were under reasonable limits. In summary, this paper established an intergrated long-term SCUC model, and developed the corresponding software.At last, although the formulation has been used to solve the SCUC problem in traditional power systems, it also could be applicable to the new scheduling problems in electricity markets.