The invention provides an intelligent optimization method for carbon emission reduction of an ecological
pasture in
extreme weather, and relates to the technical field of intelligent optimization of carbon emission reduction. Comprising the steps that initial
monitoring data are collected and preprocessed, and multi-dimensional
feature vector output state features are constructed; an LSTM
time sequence prediction model is used for analyzing the state of the dairy cow and the activity of methanogens in the
rumen, DHI data of an integrated
gradient boosting tree model is used for analyzing the
amino acid balance and precise
nutrition technology of the lactating dairy cow, a dairy cow
manure yield prediction model based on a
Bayesian network is constructed, a
manure management strategy is dynamically optimized, and
manure management effect data is output. And constructing a carbon emission reduction comprehensive evaluation model, and evaluating the
pasture carbon emission reduction effect. According to the invention, the
rumen flora structure is regulated and controlled, the
rumen greenhouse gas emission is controlled, the
ruminant feed
protein level is reduced, and the
methane emission is reduced by researching the dairy cow carbon emission rumen
microorganism regulation and control technology.