Heat tripping time is an important index of low voltage circuit breaker. In order to realize accurate heat tripping time online prediction, this paper puts forward a kind of weighted kernel partial squares (WKPLS) prediction method. Firstly, the method determines the weight value of the sample according to the distribution and statistical characteristic of the production sample data, order to solve the problem of unbalanced distribution of production data of the circuit breaker. Secondly, the heat tripping time model of low circuit breaker is established by using WKPLS method, Current, ambient temperature, initial position of bimetallic sheet and welding position as the input, and the heat tripping time is the output. Finally, a simulated experiment were conducted by using the collected circuit breaker production data, which indicate that the heat tripping time model of circuit breaker based on WKPLS method possess high prediction accuracy, and the model can provide some reference for the adjustments of control parameters and abnormal monitoring in the production process of circuit breaker.