Theoretical Study On Optimization Decision-making Of Maintenance Scheduling And Unit Commitment

Today,the depletion of fossil fuel resources and the deterioration of the ecological environment threaten the survival of humankind.Therefore,it is necessary to carry out the energy cleanup revolution and promote power system transformation and upgrading in the direction of the green,low-carbon,energy-saving,and highly efficient smart power utilization.Under the new situation,the integration of renewable energy generation in decentralized and centralized manners not only broughts energy-saving and emission-reduction but also introduces uncertainties.The uncertainties change the synchronous balance process of power generation and consumption into complex patterns of resource complementation,space-time correlation and contradicton between decentralization and centralization.As a reslut,the optimization decision-making of power system operation and control faces with serious challenges.In view of the above background,power system economic operation and control are faced with new issues that inclue how to accommodate the large-scale renewable generation,how to effectively complement a variety of energy sources and how to improve the power system efficiency in the presence of equipment condition-based maintenance.This dissertation has investigated the optimization decision-making of maintenance scheduling and unit commitment in the presence of uncerainties.The optimization decision-making models and methods for power system economic operation are studied in order to settle the contradictions appeared in the balance process of power generation and consumption.The main contributions of the dissertation are summarized as follows:(1)The basic ideas of the optimization decision-making for power system economic operation including time combination,space combination and problem combination are proposed in response to uncertainties.Based on the proposed ideas,the coordination method for maintenance scheduling,unit commitment,advanced economic dispatch,and automatic generation control is clearly defined.In addition,the collaborative strategy for transmission grid and active distribution grids is also estalished.(2)A fast serial algebraic analytical method for economic dispatch problem considering network losses is proposed.The proposed method decomposes the economic dispatch problem considering network losses varying with power generations into several economic dispatch problems with constant losses based on the convexity between losses and the power generations.Consequently the optimal solution can be found in monotone finite algebraic calculus.Meanwhile,this dissertation explores the decomposition and coordination mechanism of Lagrange relaxation technique as a basic element of the problem combination and lays a foundation for the optimization of optimal power flow,unit commitment and maintenance scheduling.(3)Variations and uncertainties in wind power generations introduce new challenge whether the unit commitment schedule is feasible to cope with constraints appeared in a time instant or in a time period.Considering this problem,intermittent uncertainties are expressed by its variation range,speed and acceleration.Then,essential conditions to accommodate intermittent uncertainties are deduced,and the power system ability to offer these conditions is evaluated.Finally,a coordinated decision-making model of unit commitment and economic dispatch is proposed.Based on the proposed model,a feasible unit commitment schedule considering wind power intermittent uncertainties is provided.Furthermore,relationship between unit commitment schedule and its ability to accommodate wind power is explored under different rigorous scenarios,together wth the coordinate benefits.(4)The proposed approach based on Benders decomposition to the optimal coordination of maintenance scheduling of generating units and transmission lines,with the security-constrained unit commitment can be used to deal with contradictions caused by the increasingly application of equipment condition-based maintenance in power system.The proposed approach decomposed the optimization problem into master problem,auxiliary problem,and power flow sub-problems,among which the dual information is transmitted via Benders cuts.In order to improve the efficiency,auxiliary problem is introduced to identify the correlation between transmission maintenance scheduling and power generation schedule.In addition,a filtering method based on pattern recognition technique is used to reduce the number of power flow sub-problems and the calculation speed is improved.(5)In the background of equipment condition-based maintenance,contradictions will be easily introduced between transmission maintenance scheduling and unit commitment,especially with a high level of uncertainty due to the wind power integration.To meet this challenge,this dissertation models the uncertain nodal demands and wind power generations through distribution-free bounded interval uncertainty set,and then proposes a robust optimization decision-making model to solve the coordination of transmission maintenance scheduling with unit commitment.To solve the proposed model,this dissertation decomposes the optimization problem into a master problem and two sub-problems based on Benders decomposition method.The master problem applies the Lagrangian relaxation technique to obtain optimal hourly results for transmission maintenance scheduling and unit commitment based on the Benders cuts that are formed in sub-problems on the basis of the feasible or optimal conditions.The sub-problems are to verify whether the decision in the master problem is feasible and optimal.Furthermore,the proposed approach introduces the minimum operating costs associated with economic dispatch problems into master problem to decrease the gap between the master problem and sub-problems for accelerating the convergence of the optimal results,and also uses the principle of duality and the characteristic of interval uncertainty set to covert the intractable max-min sub-problem models into mixed-integer linear programming models.(6)In order to release the synergistic capability of the active distribution networks with the transmission grid to accommodate uncertainty as well as to solve the contradiction between distribution and concentration,a decision-making model to the optimal coordination of transmission constrained unit commitment incorporating active distribution networks is established,in which the primary and secondary frequency characteristics are considered.To solve the proposed model,a distributed optimization approach is presented based on augmented Lagrangian method.In the solution methodology,the proposed approach takes the tie lines among the active distribution networks and the transmission grid as the decomposition and coordination points,and then decomposes the tie lines into virtual loads from the prospective of transmission grid and virtual generatiors from distribution networks point of view.According to augmented Lagrangian method,the optimization model is decomposed into transmission constrained unit commitment subproblem and several active distribution networks constrained optimization decision-making models.The parallel computing techque is used to solve the active distribution network optimization problems based on the distributed autonomous characteristics.As a result,the optimal solution under the Lagrangian multipliers given can be obtained in finite alternate iterations of transmission and distribution network constrained problems optimaization.Finally,the proposed approach applies modified Lagrangian multiplier equation to obtain feasible and optimal solutions for transmission constrained unit commitment and and active distribution constrained optimization problems.