Abstract: A method for evaluating nitrogen bio-removal capacity in water based on functional traits of benthic macroinvertebrates, comprising: S1, constructing an evaluation table for the nitrogen bio-removal capacity in water, and S2, collecting the benthic macroinvertebrates in the water to be evaluated, calculating the final functional trait evaluation scores of the benthic macroinvertebrates based on modalities that are strongly correlated with nitrogen bio-removal rate of water indicating the nitrogen bio-removal capacity in water, and then determining corresponding grade based on the evaluation table. The present application requires only the collection and identification of benthic macroinvertebrates from water to obtain the functional trait evaluation result, enabling the determination of the nitrogen bio-removal rate for the water by using the pre-constructed assessment table.

Abstract: The present application introduces a membrane fouling warning methodology grounded in machine learning. It utilizes a machine learning-based membrane fouling prediction model to automatically forecast and generate electrochemical information values, which characterize the extent of membrane fouling at various time points, based on influent water quality parameters. It then acquires the electrochemical information values Zt at a moment t and Zt+?t at a moment t+?t. Subsequently, it computes and assesses the respective fouling levels using the electrochemical information values derived from the membrane fouling prediction model. Finally, it issues an early warning signal contingent upon the determined warning level. This methodology facilitates proactive understanding and management of membrane fouling, thereby sustaining the normal operation of the membrane fouling treatment system, mitigating the propensity for membrane assembly fouling, and prolonging the operational lifespan of the membrane assembly.

Abstract: The present application introduces a membrane fouling warning methodology grounded in machine learning. It utilizes a machine learning-based membrane fouling prediction model to automatically forecast and generate electrochemical information values, which characterize the extent of membrane fouling at various time points, based on influent water quality parameters. It then acquires the electrochemical information values Zt at a moment t and Z++?t at a moment t+?t. Subsequently, it computes and assesses the respective fouling levels using the electrochemical information values derived from the membrane fouling prediction model. Finally, it issues an early warning signal contingent upon the determined warning level. This methodology facilitates proactive understanding and management of membrane fouling, thereby sustaining the normal operation of the membrane fouling treatment system, mitigating the propensity for membrane assembly fouling, and prolonging the operational lifespan of the membrane assembly.