【作者】 张坤；

【导师】 齐维贵；

【作者基本信息】 哈尔滨工业大学 ， 电气工程及其自动化， 2009， 硕士

【摘要】 集中供热逐渐成为城市尤其是北方城市建设基础之一,如何使整个集中供热系统处于一个良好的、高效的运行状态,成为供热控制系统所必须解决的问题。但供热负荷具有一定的随机性,传统的供热系统的运行仅仅是以天气情况为依据,调度人员往往根据天气情况和工作经验来调节热网运行参数,因而热网调度的精确性较差,所以对供热负荷进行预报是至关重要的。本文采用乘积季节性ARIMA模型与以往其他的调度方法进行比较,利用采集得到的供热系统的实时数据来预报下一个采样时刻的预报值,并以此为依据用于供热系统的优化调度,将对供热系统的节能运行和实时控制起到重要作用。首先对供热负荷原始数据进行预处理,将平稳时间序列作为分析的对象。接着分别利用AR横向、纵向、交叉负荷预报,乘积季节性ARIMA模型方法实现供热负荷预报,结果表明应用乘积季节性ARIMA模型进行预报的效果良好,且优于以前的预报算法。同时本文所采用的乘积季节性ARIMA模型能够进行自动完成阶数选取,使模型达到最优。论文最后将乘积季节性ARIMA方法应用于集中供热系统智能管理软件的预报调度中。系统设计分为两大部分:负荷预报和监控管理。其中监控管理主要包括数据处理系统、用户管理系统、查询系统、输入输出系统、安全报警系统;负荷预报包括预测控制系统。利用热负荷原始数据进行预报,监控管理界面将根据负荷预报值给出诸多调整建议,使操作人员可以更为直观的监控调整供热系统,指导供热网络节能运行。对于提高供热质量和节约能源都有着重要的意义。更多还原

【Abstract】 Central heating has gradually become the city one of the foundations of the northern cities. How to make the whole central heating system in a good and efficient operation become more and more necessary. However, heating load with a certain degree of randomness, the traditional operation of heating system is only based on weather conditions and work experience to regulate the heat supply network operating parameters, thus the accuracy of it is less. Therefore, on the heating load forecasting is essential. In this paper, the product of the multiple seasonal ARIMA model with the previous comparison of the scheduling method, uses the heating system has been collecting real-time data to forecast the next sampling time values. Using it as the basis for optimal scheduling of heating system, energy-efficient heating system will run and play an important role in real-time control.First of all, the dissertation adopts the original data collected in heating supply station and makes data pre-processing. Secondly, the use of horizontal, vertical, cross-load forecasting, multiple seasonal ARIMA model of the product heat load forecasting method, results showed that the application of the product of the multiple seasonal ARIMA model forecasting the effects of good and better than the previous prediction algorithm. At the same time, the multiple seasonal ARIMA model can automatically complete the selection order.Finally, the multiple seasonal ARIMA method has been applied to central heating system. The intelligent management software includes two parts; these are monitoring system and forecasting system, and monitoring system includes data processing systems, user management systems, query systems, input-output systems, security system. Monitoring and management interface will be given a lot of load forecasting value of the proposed revision, so that operators can monitor more intuitive adjustment of heating system, and guide heating system to save energy, and these are of great significance for improving the quality and energy supply.更多还原