Research Of Heating Network Control Based On Fuzzy Cognitive Maps

With the increase of the district heating supply area, it is of great significance to research on optimizing the control arithmetic of central heating supply network, and improves heating effect. Nowadays, the district heating supply has lots of problems in our country, and waste of much resource. In order to be more efficient, and the temperature can be controlled at an even level., the theme of optimizing the control arithmetic has been discussed, combining the points of "coarse control" and "fine control" during the process, puts forward to a control method which has illation ability with the techniques requirements of heating supply network.Heating supply network is a city pipe network system with complex structure and large scale. It has many characteristics, like large inertia, large hysteresis, nonlinear, strong coupling and so on, so it's difficult to construct the model exactly, and normal control methods are difficult to satisfy the requirements. This paper introduces the basic theory and the FCM (fuzzy cognitive map) arithmetic, constructs FCM model of the instrument in laboratory to achieve the control goal by FCM. FCM comes from the combination of Neural Networks and fuzzy control, expresses information by structure. FCM learns by NHL (nonlinear hebbian learning), and but normal NHL is unsupervised learning, so supervised learning NHL is found to achieves the adjustment of the illation process, besides, we combine NHL and expertise while constructing the model, making practical use of expertise to construct the FCM at first, then adjust FCM automatically by NHL, so that FCM can show the system exactly.The text constructs the control platform for heating supply network in libratory based on FCM, embeds FCM model to the control arithmetic of central heating supply network in libratory, to achieve the control method based on the model. With experiments and analysis for lots of times, it can be proved that the control arithmetic based on FCM can shorten the control time, and improve the precision and stability, so that the state of system can satisfy the requirements, and finally, the system can be controlled with an even temperature.