Fault prediction method and device, model construction method and device, processor and readable storage medium

Abstract

The embodiment of the invention provides a fault prediction method and a device, a model construction method and a device, a processor and a readable storage medium, and belongs to the field of computers. The fault prediction method comprises the following steps: obtaining parameter values of a plurality of parameters related to fault information of the equipment; and inputting the parameter values into an equipment fault prediction model to obtain a prediction result for predicting the fault type and / or the fault occurrence time of the equipment, the fault prediction model being a model obtained by performing machine learning on training data of the equipment in advance and used for outputting the prediction result. According to the method, important faults can be predicted in advance, sothat the water injection pump is maintained in advance, high maintenance cost is avoided, and good economic benefits are obtained.

Description

technical field [0001] The invention relates to a computer, in particular to a fault prediction, a model building method, a device, a processor and a readable storage medium. Background technique [0002] For equipment used in various fields, it is inevitable to fail due to various reasons after a long time of operation. After the failure occurs, it will cause certain losses, and some may even pose a threat to personal safety. When the failure occurs, it will be discovered And deal with it too late. [0003] Taking the water injection pump as an example, the water injection pump is the core power equipment for the water injection of the water well in the development of the water flooding oilfield. The water injection pump injects the water after oil-water separation into the water well to displace the crude oil in the reservoir. [0004] The water injection pumps used in the oil field are mainly divided into two types: centrifugal and reciprocating according to the working ...

AI Technical Summary

Problems solved by technology

[0007] The production failure of the water injection pump can be analyzed and diagnosed by means of manual inspection and observation, sound experience judgment, and frequency spectrum analysis based on vibration monitoring. It directly affects the efficiency of production operations, and at the same time, after multiple damage repairs, it will have a great impact on the overall life of the water injection pump

Images