为了直观地描述SCADA系统中各种参量历史、实时及进一步变化趋势，开发出以曲线形式展示SCADA系统中数据的曲线生成器，实现曲线显示、估计等功能。 曲线生成器主要包括曲线图组态、曲线视图组态和曲线估计器三部分。 曲线显示功能通过曲线图组态与曲线视图组态来实现。曲线图组态通过访问数据库及数据处理实现了SCADA系统状态量与模拟量的混合显示，在曲线图组态中还可实现对曲线编辑、曲线添加、曲线删除、曲线保存等相关功能。曲线视图组态实现曲线的灵活显示功能，把从数据库中提取的数据，以曲线形式显示于同一曲线图中，也可分别显示在不同的曲线图中。 曲线估计功能是通过曲线估计器实现的。曲线估计器应用支持向量机的方法进行曲线估计，在对不同曲线估计时，应用禁忌搜索算法对支持向量机参数进行自适应调整。曲线估计器通过历史数据对曲线估计模型进行训练，然后选择估计时间即可实现对曲线的估计，估计结果以曲线形式展示出来。曲线估计器不仅可以根据历史数据预测未来曲线，而且还可以对因自动化终端单元故障或通道障碍等原因造成曲线缺失进行估计和拟合。基于MATLAB编程语言设计了曲线估计程序，并将其封装应用于曲线生成器中，实现曲线估计功能。 根据提出的实现原理及过程，基于面向对象思想进行设计，采用Visual C++完成了曲线生成器软件的开发。最后采用电力系统实例对曲线生成器各个功能进行验证，验证曲线生成器的可行性及实用性。

To conveniently reflect the variation trend of data obtained through SCADA system, the proposed generator of the SCADA system has the functions of hybrid curve displaying and curve estimation，which can be used to monitoring the trend of parameters of SCADA. The new curve generator consists of three parts, such as the curve plot, the curve view and the curve estimator. The function of curve displaying is realized by curve configuration and curve view. The hybrid display of the analog variables and sequence events is achieved by accessing database and data processing. Several curve function is provided, such as edit, add, delete, save, and so on. Based on the curve view configuration, the data that called from the database can be displayed on one graph or more graphs. The function of curve estimation is accomplished by the curve estimator which based on the arithmetic of Support Vector Machine, as well as the self-adaptive adjustment of data with the method of Tabu-Search. After training the historical data model, the result of estimation can be displayed in the form of graph. Not only variation trend forecasting of the parameters is obtained but also a few missing data caused by terminal fault or communication channel abnormal can be obtained by the estimator. All of the processes programming are based on MATLAB, which packaged into the VC++ based curve generator, Based on above works, an application software based on VC++ 6.0, SQL Server 2000 and MATLAB is developed. Several examples are given to test the program. The results show that the proposed approach is feasible and practical.