Multiple working conditions cobalt ion concentration prediction model based on correlate relation of different working conditions judged by fuzzy correlation factor

The cobalt removal process is an important step of Zinc Hydrometallurgy, outlet cobalt ion concentration is a significant indicator to manipulate the cobalt removal process and it can provide an important reference for on-site control. However, the cobalt ion concentration cannot be detected online and the test has hysteresis which leads to the delay of the control action or miss the optimal control time. In this paper, a multiple working conditions prediction model based on correlate relation of different working conditions judged by fuzzy correlation factor is proposed to predict the outlet cobalt ion concentration of the reactor. First, a mechanism model based on the material balance and chemical reaction principles is proposed. During the test, it was found that the accuracy of its prediction was reduced when the reaction state changed. Then, considering that the cobalt removal reaction process is greatly affected by the inlet flow rate, the cobalt removal process is divided into different working conditions according to the inlet flow rate of the reactor using K-means algorithm. Finally, considering the correlation coupling of different working conditions, the fuzzy correlation factor is proposed to describe the degree of relevance between the inset flow rate and the specific model, the result obtained by fusion of different models through fuzzy correlation factors is the final prediction result. The results demonstrate the model can predict the outlet cobalt ion concentration more accurately.