The invention provides a self-adaptive multi-working-condition steel
secondary energy generation amount
dynamic prediction method, and relates to the technical field of steel energy prediction. The method comprises the following steps: acquiring
coke oven gas generation amount historical data under multiple working conditions, setting a
coke oven gas generation amount data preprocessing time interval, reading
system clock data, preprocessing the acquired
coke oven gas generation amount data, and dividing a multi-working-condition
data set; setting
particle swarm optimization method parametersand least square
support vector machine parameters, initializing the parameters, fitting
coke oven gas generation amount prediction
model parameters in
time series data by utilizing an intelligent method, and identifying working conditions to finish
coke oven gas generation amount prediction. Stable, reliable and accurate
coke oven gas dynamic multi-working-condition prediction application is realized, prediction
model parameters can be learned adaptively, and scientific data support is provided for
energy management personnel to make a gas scheduling plan, so that energy
diffusion is reduced,refined utilization is improved, production supply is stabilized, and
energy cost is reduced.