Research On Temperature Monitoring Optimization Based On Data Fusion In Biological Oxidation Pretreatment

Biological oxidation pretreatment is an effective way to improve the gold extraction rate of refractory gold deposits,which can also effectively oxidize and degrade the gold containing sulfur and minerals.In the process of industrial control,the factor of temperature has a great influence on the biological oxidation pretreatment.A reasonable control of the temperature in biological oxidation pretreatment will directly affect the activity of the oxidase of oxidized colony and the survival of the colony,which also affects the amount of gold produced in the oxidation gold field.Therefore,in the processes of temperature monitoring in actual industrial sites,there is a great significance to ensure the accurate monitoring of temperature in biological oxidation pretreatment and cover the overall of the reactor tank,which provides support for the control of temperature to guarantee the oxidation efficiency.At present,the traditional temperature monitoring method adopted in the industrial field has great limitations and defects.Due to the large volume of the pretreatment oxidation tank,the monitoring of the sensor surface can only ensure the effectiveness of the temperature measurement in a small area,and it is not effective for the oxidation tank.Therefore,in view of the problems existing in the traditional industrial field temperature monitoring of biological oxidation pretreatment-cyanide gold extraction,this paper intends to propose a temperature measurement optimization scheme to improve the monitoring accuracy while solving the above problems.The main research contents and research results of this paper are as follows:1.The biological pretreatment process,the structure and reaction mechanism of a single pretreatment oxidation tank were studied.Based on the characteristics of the pretreatment oxidation tank,a small-scale sensor network is established.Based on the sensor network structure,a multi-connected layered sensor data fusion structure is designed as the fusion framework of the data fusion algorithm;2.Based on the sensor fusion structure,a distributed data fusion algorithm based on small-scale sensor network is proposed.The algorithm includes a single sensor data processing part and a sensor network data fusion part.In the aspect of single sensor data processing,the improved EKF algorithm is introduced,and the iterative operation idea and the memory exponential weighted multivariate fading factor are introduced to improve the data processing process,which reduces the high linearity error and accumulation error easily generated by the strong nonlinear system in the data prediction process;3.In terms of sensor network convergence,the sensor-based monitoring accuracy is used as a real-time weighting criterion,and the small-scale sensor network is fused and weighted to obtain the final fusion monitoring value.The proposed algorithm effectively reduces the temperature monitoring deviation.It can be seen through numerical simulation experiments that it improves the global monitoring accuracy,makes the monitoring processing value smoother,and provides a strong basis for subsequent temperature control in practical industrial applications.