A neural network driven method for assessing the state of life of plants and animals

By applying perturbation excitation to plants and animals and combining it with multimodal sensing, and using neural networks for dynamic modeling and risk prediction, the problems of cross-species generalization and credibility in the assessment of plant and animal survival status are solved, and the accurate assessment and intelligent management of survival status are realized.

CN122154777APending Publication Date: 2026-06-05GUIZHOU PROVINCE STATE-OWNED ZAZUO FOREST FARM +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
GUIZHOU PROVINCE STATE-OWNED ZAZUO FOREST FARM
Filing Date
2026-03-05
Publication Date
2026-06-05

AI Technical Summary

Technical Problem

Existing technologies struggle to dynamically model and reliably assess the survival status of plants and animals in complex environments. In particular, they lack cross-species generalization capabilities in multi-species contexts and fail to model the coupling mechanism between external disturbances and internal states, resulting in high uncertainty in state assessment.

Method used

By applying disturbances such as light, temperature, humidity, or sound waves, and combining the response signals collected by a multimodal sensing system, state modeling and risk prediction are performed using a variable coding layer, a cross-species migration mapper, and an anomaly generator. In conjunction with a credibility assessment mechanism, early warning signals are generated.

Benefits of technology

It enables precise dynamic modeling and reliable assessment of the survival status of plants and animals, improves the ability to identify early sub-health conditions and the universality of cross-species assessment systems, and supports intelligent ecological management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122154777A_ABST
    Figure CN122154777A_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of neural network driving, and particularly relates to a kind of neural network driving's plant and animal survival state evaluation method, including the disturbance excitation signal to target plant and animal individual or the microenvironment where it is located is applied, the physiological, behavior and environmental response signal of individual is continuously collected before disturbance, in disturbance and after disturbance, time alignment and abnormal filtering are carried out, disturbance response trajectory sequence is constructed, and disturbance response trajectory sequence is input into state deduction network;Obtain high-dimensional state representation set;High-dimensional state representation set is input into classification discriminator, and according to survival state probability label, early warning grade score is generated;When early warning grade score exceeds intervention threshold, or survival state probability label is lower than credibility threshold, early warning signal is triggered to prompt management system intervention response.The present application overcomes the defect of ignoring individual dynamic response capability, and improves the identification ability of early sub-health state or implicit anomaly.
Need to check novelty before this filing date? Find Prior Art