基于状态寻优的工业系统动态数据驱动建模

Dynamic Data Driven Modeling of Industrial System Based on State Optimization

针对应用动态数据进行系统建模的过程中,系统的初始状态无法获取的问题,提出一种基于状态寻优的工业系统动态数据驱动建模方法。上述方法选取系统运行过程历史数据中的纯动态数据作为建模数据,将建模数据末端的系统的输入值作为系统输入的稳态分量,将系统输出的稳态分量、动态数据起点的系统初始状态及模型参数均作为寻优变量的维度,应用教学优化算法进行寻优,从而建立系统的模型。对某工业系统进行建模仿真,结果表明上述方法的有效性。

In order to solve the problem of obtaining the initial state of the system in the process of dynamic historical data modeling, we proposed a dynamic data driven modeling method for industrial systems based on state optimization . The pure dynamic data from the system history data were selected as the modeling data. The value of the system input at the end of the modeling data was considered as the steady-state component of system input. The steady -state component of the system output, the initial state of the system at the starting point of the dynamic data and the model parameters were taken as the dimension of the optimization variables. Finally, applying teaching-learning- based optimization algorithm to optimize the above parameters, we establish the system model. The modeling and simulation of an industrial system show the effectiveness of the method.