Abstract: Abstract: Although there are some studies on parameters inversion for hydration exothermic model without considering concrete temperature duration, there is few study on parameters inversion for hydration exothermic model considering concrete temperature duration. The adiabatic exothermic model considering temperature duration has better precision than those without consideration, but it will need more parameters in the model and make inversion harder when using finite element back analytical algorithm to determinate parameters, because back analysis of finite element method is easy to premature and difficult to solve the model with too many parameters. When using genetic algorithm to inverse parameters, the result of the parameter inversion is greatly depended on the way to complete genetic algorithm. A proper genetic algorithm method (the one with proper encode operation, proper construction of fitness function and proper crossover operation) is presented in this paper to solve this problem, and two non-adiabatic temperature rise experiments with quite different temperature duration are designed. Based on the measured data of one test block, the parameters of adiabatic temperature rise models were inversed by proposed genetic algorithm method, and then those parameters were verified with the data of another test block. The results show that the parameter inversed by one block can be well used in the other block even when temperature duration is quite different. That means it is feasible to use the proposed method for back inversion of parameters of adiabatic temperature rise models considering temperature duration. In this paper it is shown that adiabatic temperature rise is highly depended on the concrete temperature duration, the hydration exothermic will be different with the changing of temperature duration. So when without considering temperature duration, hydration model can not be expressed by one expression. By comparing the results of the parameter inversion for hydration exothermic models with and without considering concrete temperature duration, it is showed that the adiabatic exothermic model considering temperature has better precision than the one without consideration.