基于神经模糊系统的热工过程建模及预测

热工对象内部过程的物理性能比较复杂，其往往表现出非线性、严重时变、大迟延和不确定等特点，这就使得难以对其建立比较精确的模型。该文以自适应神经模糊推理系统(ANFIS)作为辨识器建立热工过程模型，用ANFIS分别建立锅炉-汽轮机的非线性模型、不同负荷工况点的线性模型，并根据现场采集的锅炉-汽轮机系统数据建立了ANFIS模型。对以上三个系统的建模仿真结果表明基于ANFIS建立的模型具有较高的模型精度和较好的预测能力，ANFIS可用于非线性系统、复杂系统的建模和预测，并具有较少的训练次数和较小的预测误差。

Modeling and Prediction of the Thermal Process Using Neuro-Fuzzy System

The physical performance of the inner process of thermal engineering objects is comparably complicated. The thermal process often has the specialities of nonlinearity, strong time - variation, long delay and uncertainty, which make it very difficult to build accurate mathematic model. In this paper, using neuro-fuzzy system (ANFIS) as an identifier for building the model of the thermal process is proposed. ANFIS is respectively used to build the nonlinear model, the linear model with different load of the boiler - turbine system, and the ANFIS model built according to the field data of the boiler - turbine system. The modeling simulation results of the three systems show that the models built by using ANFIS have higher identification precision and better predictive ability. ANFIS can be used to build the models and predict the outputs of the nonlinear and complex system. Furthermore, ANFIS needs less training time and has less predictive error.