Application of Ant Colony Optimization Least Squares Support Vector Machine in Measurement Data Fitting

Aiming at the disadvantages of traditional data fitting methods, such as relying on the user’s experience and needing to predetermine the estimated fitting function, a data fitting method based on Ant colony least squares support vector regression（ACO-LSSVR） is proposed. The method uses ant colony optimization（ACO） to optimize the parameters of least squares support vector regression machine（LSSVR） and obtain the optimal parameters to establish a data fitting model. This method is used to fit the measured data of nuclear engineering with the traditional regression fitting method. The core power curve and the melt droplet movement characteristic curve in coolant are obtained. The fitting results of the two curves are compared. Results show that ACO-LSSVR has high fitting accuracy and does not need to determine the fitting function of data segments.