The invention belongs to the technical field of intelligent driving, and particularly relates to an intelligent driving decision-making
system based on multi-
modal sensor fusion, which comprises a multi-
modal sensing layer, a data fusion layer, an operation decision-making layer and an
execution control layer, and is characterized in that the multi-
modal sensing layer comprises a
laser radar, a visual camera group, a
millimeter wave radar, a GNSS positioning module and an
agricultural machinery special sensor group; according to the method, by setting a multi-modal sensor cooperation and agricultural
scenario algorithm, in complex environments such as
crop shielding, dust and rugged
topography, the environment
perception accuracy is greater than or equal to 96%, the obstacle recognition distance is greater than or equal to 50 m, the scene limitation of a single sensor is solved, and by improving the combination of EKF and a semantic attention mechanism, multi-
source data space-
time alignment and weight optimization are realized; the
signal-to-
noise ratio of the fused data is increased by more than or equal to 15dB, high-
quality information support is provided for
decision making, operation parameters can be dynamically adjusted according to
crop types and soil conditions by setting a mixed decision-making framework, the path planning
overlap ratio error is less than or equal to 3%, and the operation quality deviation is controlled within + / -5%.