摘要
针对电厂燃煤管理系统的运行特点,提出了适合电厂运行的在线配煤优化模型。采用神经网技术对混煤的煤质参数进行预测,用改进的遗传算法对配煤方案进行求解。仿真实验证明,神经网络技术解决了加权平均法和经验公式法预测精度不高的问题。用遗传算法得到了符合生产实际的配煤问题的最优方案,确保了实时计算效率,有一定的工程实用价值。
According to the characteristics of the coal management system of power plants, an online coal blending optimization model is established for plant operation. Mixed coal parameters are predicted by neural network. The coal blending problems are solved by improved genetic algorithm. Through analysis of simulation, neural network the defects of low predictable accuracy of the weighted average method and empirical formula are overcome. The optimal solution for coal blending is obtained with the genetic algorithm, which is consistent with the actual production situation. The coal blending model has real-time performance and practical value.
出处
《现代电力》
2008年第1期67-72,共6页
Modern Electric Power
关键词
燃煤管理
配煤
神经网络
遗传算法
优化
coal management
coal blending model
neural network
genetic algorithm
optimization
