Scrap disposal strategy actuarial method based on information feature combination
By uniformly processing and combining features of multi-source data on end-of-life materials, a revenue-risk actuarial model is constructed, which solves the problems of inconsistency in multi-source data and insufficient feature correlation, and realizes scientific decision-making and economic improvement in the disposal of end-of-life materials.
CN122243220APending Publication Date: 2026-06-19SHANDONG LUNENG PROPERTY CO
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANDONG LUNENG PROPERTY CO
- Filing Date
- 2026-05-15
- Publication Date
- 2026-06-19
Smart Images

Figure CN122243220A_ABST
Abstract
This invention discloses an actuarial method for the disposal strategy of waste materials based on information feature combination, belonging to the fields of data processing and intelligent decision-making technology. The method includes: constructing multi-source waste material information features; performing feature combination modeling based on multi-dimensional information features, modifying feature weights and forming a structure-preserving combined feature expression by introducing a credibility assessment mechanism based on source and time sequence; constructing a disposal benefit-risk actuarial model, quantifying the recovery value, reuse value, and dismantling value of different disposal paths, and comprehensively evaluating them in conjunction with transportation risk, price fluctuation risk, and time delay impact; and performing disposal decision optimization based on the actuarial results, generating a disposal strategy under resource constraints, and outputting corresponding execution instructions. This invention can achieve refined quantitative evaluation and intelligent optimization decision-making for waste material disposal paths, improve disposal benefits and risk control levels, and has good engineering application value.
Need to check novelty before this filing date? Find Prior Art