Research On The Dynamic Data-driven For Predicting Air Input Of Bio-oxidation Tank

As an important process of gold extraction,biological oxidation pretreatment process determines the extraction rate of gold.Given the intake air oxidation tank in the biological pretreatment of the value directly affects the efficiency of oxidation tank activity and bacterial oxidation of refractory gold ore,and finally affect the gold extraction rate,and air intake system of biological oxidation pretreatment process is the electric power consumption is one of the main system of the whole process for biological oxidation.At present,given the air intake is an open loop system,generally by the experienced workers to manually adjust the gas pipeline valve,air intake and control the oxidation tank,which uses a method more rather not less,resulting in high dissolved oxygen levels and oxygen utilization,resulting in waste of energy.Secondly,in the process of biological oxidation pretreatment,because of the influence of slurry stirring flow,pH value,temperature change of oxidation tank and bacterial oxidation activity,the dynamic and uncertainty of air intake system is increased.Therefore,in the open loop air intake system of the biological oxidation tank regulation,energy waste and uncertainty problems,put forward a kind of biological oxidation tank inlet dynamic data driven based on accurate prediction method,so as to construct the closed-loop air intake control system provides the basis for accurate.The main contents of the research are as follows:(1)first,is predicted by the model simulation of biological oxidation tank air intake system is established,with multistage biological oxidation tank system as the background,the state space model,the establishment of air intake on the single oxidation tank and secondly,by taking into account the correlation between the levels of oxidation tank inlet amount and the coupling of sequence data to establish the oxidation tank the gas into the state space model by multistage oxidation tank gas input time.(2)finally,the data assimilation method based on Kalman algorithm is applied to realize the dynamic integration of prediction data and measured data,and real-time and dynamic update of prediction values and model parameters.In order to verify the validity and accuracy of this method,the simulation based on real data of a Xinjiang cyanidation plant,the experimental results show that the prediction framework can be accurately predicted,the biological oxidation tank and air intake,considering the prediction of the state space model to establish a multi-dimensional correlation under more accurate results.