Prediction Of PH Value During Biological Oxidation Pretreatment

The biological oxidation gold extraction process is increasingly valued by the ore smelting industry because of its low investment cost,green environmental protection,and the advantages of meeting the current national green development.The biological oxidation pretreatment process is the core link of the biological oxidation gold extraction process.This process uses ore-leaching bacteria to decompose the associated mineral impurities containing sulfur,arsenic and iron in the difficult-to-treat gold ore,thereby precipitating gold.The pH value is a key factor affecting the activity and growth rate of ore-leaching bacteria,and the activity and growth rate of bacteria determine the oxidation rate and ultimate gold extraction efficiency of refractory ores.Therefore,predicting the pH value in the oxidation tank is of great significance to the improvement of the biological oxidation pretreatment process and the improvement of production efficiency.The main research contents of this article are as follows:According to the analysis of the process flow in actual production,combined with the mechanism of oxidation of the ore leaching bacteria,the data collected from the actual process site was used to analyze and obtain the optimal pH range of the slurry,which provides a reference for the study of pH value prediction.For the establishment of the pH prediction model during the biological oxidation pretreatment process,the pH in the oxidation tank has the characteristics of hysteresis,time variability,dynamics and nonlinearity,plus the complicated working conditions and the noise caused by the pH value acquisition equipment,making the actual The measured pH data has great volatility and randomness;for the problem of volatility and randomness of the original pH data,this paper uses a wavelet analysis denoising data preprocessing method to improve the accuracy of pH data;To deal with the time-varying and dynamic problems of the process,this paper improves the least square support vector machine algorithm(LSSVR),and establishes an improved LSSVR prediction model with data update and feedback correction capabilities.In order to further improve the accuracy of pH prediction,according to the characteristics of particle swarm optimization and hybrid leapfrogging algorithm,the two algorithms are fused to obtain hybrid leapfrogging particle swarm optimization(PSO-SFLA)for optimizing the parameters of improved LSSVR prediction model.According to the established pH prediction model and parameter optimization algorithm,a real-time pH prediction system is designed,and the workflow of the real-time pH prediction system is summarized.By comparing with the improved LSSVR prediction model optimized by the ordinary particle swarm optimization algorithm and the improved LSSVR prediction model,through simulation and comparative analysis,it can be seen that the pH prediction method proposed in this paper is feasible and effective.