Control method based on Fuzzy cognitive map and its application on district heating network

As we all know, many complex industrial processes are difficult to establish mathematical models and control. Fuzzy cognitive map (FCM) is a very convenient, simple, and powerful tool for simulation and analysis of dynamic systems. However, the construction of FCM relies on experts´ experience, which limits the application of FCM in the complex industrial process control. This paper proposed a method to establish FCM model of controlled objects based on historical data and using least square. We also propose a method utilizing δ learning rules to make FCM imitate human reasoning ability to realize control of the complex industrial process. As for the district heating network model, we use this method into realizing even heat supply. The results of simulation show that this method has the features like simple algorithm, small number of iteration steps and good robustness, it will be suitable for real time control of the complex industrial process.