The uncertainty of transport environment threatens the comfort and health of live mutton sheep and even affects the quality of meat after slaughter. Using modern information technology to solve the deficiencies of live animal centralized transport is of great significance for the stable development of animal husbandry. In this paper, the wearable multi-sensor system was specially designed, and the system test experiment and transport monitoring experiment were carried out. According to the continuous and real-time data of environmental and physiological parameters obtained from the transport monitoring experiment, the optimization extraction, data collection analysis, correlation analysis and simulation analysis of the internal environment in the carriage were carried out, and the prediction model between environmental and physiological parameters based on generalized regression neural network (GRNN) and the prediction model of comfort and health evaluation based on back propagation neural network (BPNN) were established. The results show that: (1) The wearable multi-sensor system had high accuracy and stability of data collection, and the power consumption and communication performance can meet the monitoring requirements. (2) The observation values of internal environmental parameters increased gradually with the accumulation of transport time, and finally reached the state of dynamic balance. The heart rate fluctuated greatly and was higher than the normal range. The blood oxygen saturation showed a gradual decrease trend, but the overall was in the normal range. The body temperature gradually increased, and was affected by discomfort. (3) The correlation analysis showed that there was a significant correlation between environmental and physiological parameters. (4) The area with larger value of environmental parameters was close to the middle and front of the carriage, which was mainly caused by the high density and poor air circulation. (5) The prediction model between environmental and physiological parameters based on GRNN and the prediction model of comfort and health evaluation based on BPNN had high prediction accuracy, and the combination of the two models can map the impact of environmental factors on the change of vital signs to the relationship between comfort and health. Therefore, in the case of unknown vital signs information, only obtaining environmental information can also predict the health levels, which can provide decision-making basis for relevant practitioners.
Read full abstract