In order to establish a thermal error prediction model with high prediction accuracy and generalization performance

a thermal error pseudo-hysteresis prediction model based on metabolic theory is proposed in this paper. The pseudo-hysteresis effect of machine tool is found by experimental research

and it is assumed that the thermal error is the result of the coupled action of the temperature rise at key points and the thermal error of the previous moment

and the thermal key points of the machine tool and the average lag time in the typical working conditions are solved. The genetic algorithm is used to optimize the structural parameters of the least squares support vector machine (LS-SVM). Based on the principle of metabolism

the thermal error is iteratively solved and the thermal error pseudo-hysteresis prediction model of the machine tool is established. The results of different prediction models show that the correctness of the hypothesis and the prediction accuracy of the pseudo-hysteresis prediction model is higher

the generalization performance is better

and the thermal error of different rotational speeds can be reduced by more than 90%.