Abrasion prediction optimization method for boiler water wall system

Abstract

The invention relates to the technical field of coal-fired boiler application, and particularly provides a boiler water-cooled wall system wear prediction optimization method, which comprises the following steps: carrying out gridding partitioning on a water-cooled wall system of a boiler to obtain a plurality of detection regions, and obtaining wear historical data corresponding to the plurality of detection regions respectively; constructing a wear prediction model based on the wear historical data, wherein the wear prediction model is used for representing a corresponding relationship between the wear historical data and monitoring operation parameters of a plurality of historical detection times in the wear historical data; based on the monitoring operation parameters, establishing a numerical simulation model of the water-cooled wall system, and generating wear evaluation data corresponding to the target wear area; and predicting the current wall thickness data of the target wear area by adopting the wear prediction model, and optimizing the monitoring operation parameters under the condition that the difference value between the wear evaluation data and the current wall thickness data is greater than the preset threshold value, so that the predicted wear can be continuously optimized under the condition of considering the monitoring operation parameters in the hearth, and the prediction accuracy is improved. Therefore, the wear prediction accuracy is improved.