Research On Moisture Control Of Sintering Return Powder Based On Fuzzy Neural Network

The sintering process is an essential part of the Imperial Smelting Process.While producing qualified products,the products that fail to meet the standard(sintering return powder)need to be recycled by adding water.In order to avoid dust and sliding,the moisture of the returned powder must be stabilized within the expected range during the recovery process.It is difficult for the traditional manual water adding method to achieve the ideal control effect.The continuous development of intelligent control technology provides a new idea for the process of adding water to return powder.In this paper,the technological process of lead-zinc ore sintering is analyzed.The characteristics of water feeding process and the difficulty of water control are summarized.Aiming at the problem that it is difficult to precisely model the water addding process,the fuzzy control method is adopted to realize the automatic water addition of the process.Firstly,using the traditional fuzzy control method,a Mandani fuzzy controller with the input of feed frequency and feed temperature is designed,which basically realizes the automatic control of the moisture of the returned powder.The moisture control effect is qualitatively improved compared with the manual method,but the control accuracy has not reached the expected goal.Then,according to the characteristics of multi factor disturbance in return powder moisture control,a decision-making oriented compensation fuzzy neuron is introduced,and a return powder moisture controller based on multi input single output compensation fuzzy neural network is designed.combining expert experience with the selflearning function of neural network,the ability of neural network to modify parameters and membership functions online is combined with the ability of fuzzy system language knowledge expression.According to the characteristics of multi factor disturbance in return powder moisture control,a decision-making oriented compensation fuzzy neuron is introduced into the fuzzy neural network model,which enhances the fuzzy reasoning ability of neural network model,The precision of moisture control of returned powder is improved.The experimental results show that the moisture controller based on compensated fuzzy neural network can stabilize the moisture content of returned powder within ± 1%,and the control effect is significantly better than that of manual water addition and control.The automatic control system of returned powder moisture,which is composed of PLC,touch screen,industrial computer and other equipment,has good operation performance.The water value of returned powder is stable within the expected range,avoiding dust and sliding,improving sintering efficiency and protecting the ecological environment.Figure 51,Table 9,Reference 68...