Abstract: A monitoring system of agricultural greenhouse based on Multi-data fusion +BP neural network was studied. A Multi-point sensor was set inside the greenhouse to avoid the problem of inaccurate measurement of a single sensor. The weighted average algorithm was used for data fusion, and the BP neural network was combined to judge the changing trend of parameters. Finally, the control strategy after decision was obtained. This study had achieved a more accurate prediction of greenhouse environmental parameters, and the parameters after Multi-data fusion were more accurate. Then, neural network training was carried out to obtain the trend of parameter changes, so as to provide a good plant growth environment for agricultural greenhouse.

Key words: greenhouse, BP neural network, Multi-data fusion, weighted average