The invention provides an intelligent meteorological parameter fusion processing method for recognizing floors by a carrier robot. According to the method, a data is collected by a temperature sensor, a humidity sensor and an air pressure sensor with relatively low cost so as to establish a database; aiming at the characteristic that the air pressure changes along with the height, a prediction model is establishing by utilizing an SFLA optimized MKSVM model by adoption of a big data processing technology, so as to realize the adaptive elevator floor identification of the carrier robot under various environments, and then effectively solving the condition of incorrect floor identification caused by changes of air pressure values at a same height under different environments. Moreover, the inner parts or outer parts of elevators do not need to be transformed, so that the method has extremely high universality.