Equipment Health Index Prediction Model Based On Emergency Production Plan

Insert a single phenomenon has been prevalent in the workshop. To be able to complete a single plug production tasks, companies often develop new production plan. Changes in the new production plan means that the device work load. This change will directly affect the decline rate of equipment health indicators. The change of equipment preventive maintenance time will eventually be caused. This often leads to equipment maintenance lag or excessive maintenance, resulting in economic losses.In this paper, the characteristics of equipment health index and the rate of job load change are analyzed deeply. The relationship between dynamic production plan and health index was established. Mainly do the following work:1. According to the theory of constraints, the bottleneck resources are put forward. Considering other influencing factors of order priority, the weight of influencing factors of order synthesis is calculated. A method for determining the priority of interpolation is obtained. Using the method of inserting the plug on the extended list scheduling determine the new production sequence.2. Based on the principle of gradient descent method, the health index curve is divided, and the acceleration of health index is calculated. The effects of the device’s own decline factor and environmental factor are excluded. Based on the linear regression equation, the function relationship between the acceleration of equipment health and the rate of load change is established.3. By dividing the interval of the line graph of the health index,determining the rate of load change, Calculating device health index decline rate, Cumulative health index decline, and Comparing the value of residual maintenance life, a model of equipment health index prediction based on dynamic load is established. Predictive value of health index prediction model based on dynamic job load is batter than the predictive value of health indicators of degraded hidden semi Markov model. It is proved thatthe health index prediction model established in this paper is feasible and reliable.This paper presents a method based on the bottleneck resource efficiency, which is based on the interpolation method. It is a new method to determine the interpolation order. The influence of dynamic load on health indicators is solved. Compared with other methods, it is more accurate to predict the time of equipment maintenance. Reduce the cost of maintenance lag and excessive maintenance.