Online Economic Dispatch Based On Robust Optimization Method

Wind power, photovoltaic and other new energy power which are renewable and non-pollution, are useful to relieve the shortage of energy. However, affected by climate, environment and other factors, the new energy power generation has obvious randomness and intermittency. Large scale new energy power integration will increase the uncertainty of power system and bring hidden danger to the power system operation. Thus, there is much significance on the study of uncertainty dispatch.Many kinds of uncertainty optimization methods, such as stochastic programming, fuzzy programming and robust optimization, are used in real-time dispatch of power system. Based on probability distribution of uncertain variables, probabilistically optimal solutions can be obtained by stochastic programming, after considering all random scenarios. However, the difficulties in obtaining accurate probabilities distribution functions and computational complexity limit its application. Fuzzy programming uses subordinating degree function to reflect the attitude of the decision-maker of uncertainty variables and their influence. By maximizing the membership values, optimal solutions can be obtained. However, the solutions are affected obviously by the subordinating degree function, which will lead to subjective solutions. Unlike the above two methods, robust optimization needs no probabilities distribution functions and the solutions have no subjectivity of the decision-maker. Instead, RO can capture robust optimization results with respect to a given disturbance range by satisfying the worst case within the range.Currently, there are several following issues of robust real-time dispatch:① As the fast development of new energy resources, depressing capacity of the controllable units cannot matching the new energy integration. Fully accepted model does not conform to the reality. ② Real-time dispatch model are mostly focusing on security, which will make it more conservative. Though many studies on improving conservative level of the model are done, there are still promotion spaces. ③ Most models use direct-current flow to express transmission constraints, however, the true injection shift factors has more uncertainties compared with direct-current flow model. This part of uncertainties is usually ignored.Firstly, this paper proposed a real-time dispatch bi-objective model based on effective steady-state security regions of power system, in which the base points and the participation factors are taken as control variables. Economic dispatch results can be obtained without degrading the operating security level of the power system. Secondly, Conservative level controllable robust online economic dispatch model is proposed, by adding control parameter β to improving preemptive goal programming methods. By this means the tradeoff between the operating security and economics can be tuned, at the same time, the solutions of the model are Pareto optimal. Finally, robust real-time dispatch model considering uncertainties of injection shift factors is proposed. The objective function minimizing operation cost. By using measurement data to calculate injection shift factor intervals to express transmission constraints, the robust online economic dispatch model considering both injection uncertainties and injection shift uncertainties is build. The solution can accord well with operation situation of power system.