中文摘要：

根据发电厂商的风险偏好，提出了基于效用函数的机组检修规划模型，用以规避市场电价预测不确定性及机组突然故障带来的风险。首先，分析了影响发电厂商经济效益的相关不确定性因素，采用MonteCado法及SBR（Simuhaneous Backward Reduction）降阶技术模拟可能存在的场景以评估其影响水平。然后，基于日前电能市场交易模式及发电厂商面临的风险．分析了规划期内的经济效益，主要包括各时段的售电利润、更新费用及检修费用，并根据发电厂商的自身风险偏好，构建了基于期望一风险的检修效用函数，在其经济效益期望与风险之间取得平衡。基于此，以规划期内发电厂商的效用函数最大为目标，考虑相关约束后，利用遗传算法和线性规划方法确定各机组检修时段，并用简单算例进行了定量验证。与常规模型相比，本文充分考虑了不确定性因素及发电厂商的风险偏好对检修规划、经济效益及机组出力的影响，规避了相关风险损失。

英文摘要：

This paper proposes a novel unit maintenance scheduling （UMS） model for a power producer to maximize its potential utility in the entire scheduling horizon, considering the corresponding uncertainty factors of energy prices and unexpected unit failures as well as its risk-preference attitude. Firstly, the producer＇s payoff is analyzed in detail according to the energy market, mainly including the energy-selling profits, renewal costs associated with unexpected failures and the maintenance costs. Based on this, the producer＇s utility function is investigated to achieve a tradeoff between its expected benefits and the related risk and then the UMS issue could be formulated to optimize the producer＇s potential utility within the whole periods, which is solved by the Genetic Algorithm （GA） and the linear programming methods. In the proposed model, the Monte Carlo approach and SBR （simultaneous backward reduction） technique is applied to simulate the possible scenarios of energy price and unit operating state occurring at a certain probability. Compared with the existed models, the proposed UMS model considers the influences of the related uncertain factors on the outage scheduling, economic benefits and power outputs, and lowers the power producer＇s corresponding risks. Finally, a producer with 4 generating units is utilized to illustrate its usefulness in practical application.