Conventional real-time optimization (RTO) requires detailed process models, which may be challenging or expensive to obtain. Data-driven correlation integral based RTO method is an attractive alternative to circumvent the challenge of developing accurate process models. However, the searching step size in conventional correlation integral based RTO is set by trial-and-error. In order to improve the efficiency of the correlation integral based real-time RTO method optimiszation algorithm and enhance the applicability of the system, an Bayesian optimal optimisation step-size control strategy using Bayesian optimization based on correlation integral is proposed for the traditional correlation integral optimisation method applied in industrial systems. Based on the data-driven steady state model, the adaptive control of the step size is achieved by avoiding the change of the tuning step size through the trial-and-error method due to different working conditions during the real-time optimisation control. Based on the proposed method, the application software has been developed. The simulation and industrial application results have verified the feasibility and effectiveness of the proposed method.original real-time optimisation software is improved, and the practicality of the method is proved by simulation and industrial practical application.