Taxi order demand prediction is of tremendous importance for continuous upgrading of an intelligent transportation system to realise city-scale and personalised services. An accurate short-term taxi demand prediction model in both spatial and temporal relations can assist a city pre-allocate its resources and facilitate city-scale taxi operation management in a megacity. To address problems similar to the above, in this study, we proposed a multi-zone order demand prediction model to predict short-term taxi order demand in different zones at city-scale. A two-step methodology was developed, including order zone division and multi-zone order prediction. For the zone division step, the K-means++ spatial clustering algorithm was used, and its parameter k was estimated by the between-within proportion index. For the prediction step, six methods (backpropagation neural network, support vector regression, random forest, average fusion-based method, weighted fusion-based method, and k-nearest neighbour fusion-based method) were used for comparison. To demonstrate the performance, three multi-zone weighted accuracy indictors were proposed to evaluate the order prediction ability at city-scale. These models were implemented and validated on real-world taxi order demand data from a three-month consecutive collection in Shenzhen, China. Experiment on the city-scale taxi demand data demonstrated the superior prediction performance of the multi-zone order demand prediction model with the k-nearest neighbour fusion-based method based on the proposed accuracy indicator.

In the event of a wildfire, government agencies need to make quick, well-informed decisions to safely evacuate people. Small communities, such as in Marin County, with a mix of residences and flammable vegetation in Wildland-Urban Interface zones tend to lack resources to conduct evacuation studies. Consequently, this study uses a framework of wildfire and traffic simulations to test the performance of potential evacuation strategies, including reducing the volume of evacuating vehicles through car-pooling, phasing evacuations by staggering evacuation times by zone, and prohibiting street parking in four representative areas of Marin County. Results show that reducing vehicle numbers lowers the average travel time by 20%-70% and average exposure time to wildfire by 27%-60% from the baseline. Phased evacuations with suitable time intervals lower the average travel time by 13.5%-70%, but may expose more vehicles to fire in some situations. Prohibiting street parking yields varying results due to different numbers of exits and evacuees. In some cases, prohibiting street parking reduces the average travel time by over 50%, while in other cases it only reduces the average travel time by 9%, contributing little to evacuation efficiency. Altogether, Marin County may want to consider developing a communication and parking plan to reduce the number of evacuating vehicles in wildfire situations. Phased evacuation is also highly recommended, but the suitable phasing interval depends on the speed of fire spread and number of evacuees. Further, whether to establish street parking prohibition policies for a certain area depends on the number of exits and the number of vehicles on the streets.

Many small, resource-strapped communities located in areas vulnerable to wildfire don’t have resources to conduct dedicated evacuation studies and many do not consider the impact of background traffic (i.e., normal traffic rather than evacuating traffic) on evacuation. In response, we explored the performance of several generalizable evacuation strategies with background traffic for representative communities in Marin County, including the Ross Valley, Woodacre Bowl, Tamalpais Valley, and an area near Highway 101 and Ignacio Boulevard in Novato (hereafter referred to as ‘Novato Neighborhood’). The strategies we explored include vehicle reduction (i.e., evacuees share a vehicle), phased evacuation (i.e., evacuees in different zones have different departure times), and off-street parking (i.e., street parking is prohibited on a high-fire Red Flag Day to increase overall road capacity in the event of an evacuation). We then tested each strategy using a wildfire-traffic simulation framework.