Search

The cultivated strawberry Fragaria ananassa Duch. is widely planted in greenhouses in China. Its production heavily depends on pollination services. Compared with artificial pollination, bee pollination can significantly improve fruit quality and save considerable labor requirement. Multiple factors such as bee foraging behavior, planting pattern and the spatial complexity of the greenhouse environment interacting over time and space are major obstacles to understanding of bee pollination dynamics. We propose a spatially-explicit agent-based simulation model which allows users to explore how various factors including bee foraging behavior and strawberry phenology conditions as well as the greenhouse environment influence pollination efficiency and fruit quality. Simulation experiments allowed us to compare pollination efficiencies in different conditions. Especially, the cause of bee pollination advantage, optimal bee density and bee hive location were discussed based on sensitivity analysis. In addition, simulation results provide some insights for strawberry planting in a greenhouse. The firmly validated open-source model is a useful tool for hypothesis testing and theory development for strawberry pollination research.

Indoor environment is a closed and narrow space and may exist many obstacles. The radio signal engenders multipath effect and non-line-of-sight (NLOS) propagation phenomena in indoor communication caused by refraction and reflection due to interference of the spatial radio wave signal, which leads to multipath interference and non-line-of-sight error thus consequently increases indoor positioning error. To solve these problems, based on the fact that performance of error is different from the useful signal in the wavelet domain, we proposed an adaptive wavelet de-noising algorithm (AWDA) for ultra-wideband (UWB) indoor location in this paper. The algorithm decomposes the distance value transformed from the time difference of arrival (TDOA) measured value with wavelet, then sets the threshold to detail coefficient obtained from the decomposition and cuts off error beyond the threshold. The advantage of the algorithm is that it does not need to set the threshold manually and adopts the back propagation (BP) neural network to training out the threshold which is expected to be suitable for different channels. The validity and adaptability of the proposed method had been tested and evaluated by simulation experiments using Matlab. Performance evaluation was based on comparison between adaptive threshold training and using the soft threshold or the hard threshold model in wavelet analysis. Simulation results demonstrated that the proposed algorithm has significantly reduced the NLOS error and improved the accuracy as well as the performance of indoor location under the non-line-of-sight environment.

We presented a mathematical model of population dynamics among a Model species and its Batesian Mimic as well as another species which signal is distinguishable to the Model-Mimic system. The three species are foraged by a pollinator species based on its search image in which preference can be updated by evaluating tradeoff between reward and expectation. The population dynamics can be manipulated by three keys factors: the satisfaction of reward offered by Model, floral similarity between Model and Mimic, and pollinator's learning ability. Using dynamical systems theory, we analyzed the local stabilities of the model. Moreover, we derived conditions under which the system is globally asymptotically stable from results of simulation. By using numerical simulations, the model allows scientists to evaluate the indirect effects of changes in Model-Mimic system on other coexisting species.