In order to solve the problem that the temperature sensor is susceptible to humidity during measurement,this paper takes some improvement researches into advolution operation,alienation operation and convergence conditions in mind evolutionary algorithm,and makes some adaptive adjustment of the distributing weights in the convergence operation.Differential operation is introduced into alienation operation and the situation with flat searching plane is considered.A humidity compensation method using improved mind evolutionary algorithm of BP neural network model is established,according to the sample data obtained from the experiment of humidity influence.Compared with general BP neural network model,the improved mind evolutionary algorithm of BP neural network model has higher compensation precision,faster convergence and less computation load,which can improve the accuracy and reliability of the temperature sensor effectively and is convenient for applications.