Hydrometallurgy of zinc is the world's mainstream zinc smelting method,and purification is an intermediate process in the hydrometallurgy of zinc process,which primary function is to remove impurity ions such as copper,cobalt,cadmium,and germanium from zinc sulfate solution.In order to address the challenge of low real-time monitoring values and an excessive amount of zinc powder add-ed in the purification and cobalt removal process,an oxidation reduction potential-based hybrid intel-ligent control scheme was proposed.Firstly,the potentiometers were installed in the purification tanks to obtain the real-time oxidation reduction potential that could reflect the change in the production en-vironment.Based on this,a hybrid intelligent control framework combining three modules,including potential optimization setting,potential compensation adjustment,and potential stabilization control,was designed to develop an automatic control system and applied to the actual production process.The potential optimization module utilizes artificial neural network to establish the mapping relation-ship between the input and output variables of the purification tank.Then it applies a particle swarm algorithm to optimize the potential settings of different purification tanks.The potential compensation adjustment module employs fuzzy rules and case-based reasoning to achieve real-time adjustment of potential settings during the production process.Finally,the potential stabilization control module uses a three-dimensional fuzzy rule table to adjust the amount of zinc powder added and stabilize the po-tential inside the purification tank at the set value.The results of the control experiment indicate that the proposed hybrid intelligent control framework effectively reduces the amount of zinc powder added by 4.9％while ensuring that the cobalt ion concentration at the outlet of the purification process meets the required standards,thus providing an excellent practical application value.
cobalt ion concentrationoxidation reduction potentialhybrid intelligent controlfuzzy controlcase-based reasoning
NETL
NSTL
万方数据
维普
