obtaining gas monitoring data based on a data acquisition instruction, and determining whether the gas leakage occurs;

in response to a determination that the gas leakage occurs, generating a control instruction based on a fan operation strategy to control operation of a fan, and obtaining the gas monitoring data under the fan operation strategy, the fan being configured to remove leaking gas, and a type of the gas leakage including a continuous leakage and an occasional leakage;

generating a notification instruction based on the gas monitoring data in conjunction with an operating strategy of an indicator light, and controlling the indicator light to issue an alarm notification based on the notification instruction; and

determining the type of the gas leakage by processing the fan operation strategy, the operating condition of the indicator light, and the gas monitoring data of at least one time point through a type determination model, wherein the type determination model is a machine learning model, the type determination model includes an extraction layer and a determination layer, the extraction layer is configured to determine gas concentration change characteristics by processing the gas monitoring data of the at least one time point, and the determination layer is configured to determine the type of the gas leakage by processing the fan operation strategy, the operating condition of the indicator light, and the gas concentration change characteristics.