136879 research outputs found
Sort by
The Psyche of Policing and Policing the Psyche: Conceptualizing and Managing Fatal Conflicts Between Police Officers and Seriously Mentally Ill Individuals
Individuals with serious mental illness (SMI) experience a disproportionate rate of violence and incarceration in their interactions with law enforcement officials. Unfortunately, the training method most frequently used to address this issue, Crisis Intervention Team (CIT) Training, has been shown to be ineffective in promoting behavior change in law enforcement officials. Using the most recent evidence-based health communication and conflict resolution literature, as well as a novel conceptualization of the unique forms of conflict that arise between law enforcement and individuals with SMI, this study evaluates federal standards for the delivery of CIT curriculum and offers recommendations as to how this curriculum could be improved
Development of a Framework for Safe Operation of Automated Vehicles Using Model Predictive Control With Consideration of Roadway Geometry
Safety is a crucial component of transportation design and operations. Practitioners utilize various references to ensure that roadways meet safety, operational, and sustainability requirements. Nevertheless, human error remains as a contributing factor toward unsafe driving behavior and potential crashes. In particular, rural horizontal curves have proven particularly challenging and have been the sites to a large number of fatalities.
Connected and automated vehicles (CAVs) have demonstrated the potential to enhance traffic safety and operations. Although sensor perception ranges and capabilities pose challenges, the sharing of information of information via Vehicle-to-Everything (V2X) communication has proven to be a potential solution for overcoming sensor limitations. However, due the present lack of commercially available CAVs or relevant data, there exists a gap of knowledge regarding how CAVs may operate on two-lane highways. Notably, low CAV volumes may result in behavior that can be considered analogous to that of simple, non-connected automated vehicles (AVs).
Numerous past works have turned to microscopic traffic simulation to predict CAV and AV behavior in anticipation of their growth in numbers on public roads. While the use of microscopic simulation to gain insight regarding the impacts of CAVs in urban areas is well documented by now, there is a limited knowledgebase concerning the simulation of CAVs in rural environments. This poses a problem for researchers and practitioners alike when trying to make predictions regarding CAV and AV operation and safety in such locations. The general inability of microscopic simulation to capture detailed lateral motion behaviors exacerbates the uncertainty in predicting CAV and AV behavior. Furthermore, many potential control systems exist for assisting CAVs and AVs navigate through challenging locations while considering safety constraints but most relevant schemes have been limited to varying degrees.
This dissertation attempts to address each of the aforementioned concerns. Firstly, a framework for simulating longitudinal behavior of CAVs and AVs is presented and analyzed. Second, a sensitivity analysis of lateral behavior in the context of microscopic simulation is conducted to ascertain the relationship between longitudinal speed, curve radius, and lateral motion characteristics. Lastly, based on the findings obtained from examining the aforementioned concerns, a Model Predictive Controller (MPC) working in tandem with an alignment classification heuristic is developed and assessed within alignments of varying length and curve radii.
Several notable findings were made based on the work presented in this dissertation. Firstly, V2X sensor sharing can provide significant benefits to CAV performance and reduce safety risk. On the other hand, AVs with limited sight distance demonstrate greater risk at rural horizontal curves. Secondly, the sensitivity analysis illuminates the degree to which lateral jerk and cross-track error are inversely related. Lastly, the proposed MPC scheme improves the lateral safety performance of simulated isolated CAVs, offers a viable transition path in the absence of physical spiral transitions, and demonstrates the ability to adapt to long, dynamic environments
Irrigation Water Quality, Critical Salt Levels for Peanuts, Cotton, Corn & Grain Sorghum
Multi-Fidelity Simulations of Hypersonic Turbulent Boundary Layers in the Presence of High Reynolds Numbers and Cool Walls
Hypersonic flows present many challenges that stand in the way of designing practical flight systems. In an effort to advance the state of the art in hypersonic aerodynamics, this dissertation targeted hypersonic turbulent boundary layers on a conic ogive geometry. Specifically, the research was concerned with the effects cool walls and high Reynolds numbers have on a hypersonic turbulent boundary layer and how modern turbulence modeling tools are able to perform in these circumstances. Multiple computational turbulence modeling methods were examined. At the lowest end of the modeling fidelity spectrum tested were Reynolds Averaged Navier Stokes (RANS). The RANS models tested were the Baldwin-Lomax, Spalart-Allmaras, and Menter���s SST models. The medium fidelity model was an equilibrium style Wall-Model Large Eddy Simulation. Lastly, the highest fidelity turbulence dataset generated was a 3.1 billion cell Direct Numerical Simulation (DNS) which acted as a source of validation data. The RANS models had mixed performance; Baldwin-Lomax and Spalart-Allmaras clustered together and provided reasonable comparisons to the DNS, while SST introduced significantly more uncertainty. The Wall Modeled Large Eddy Simulation (WMLES) was easily superior to RANS for all tested quantities except in predicting mean wall density. Further, the WMLES was able to reasonably predict turbulent statistics (outside the wall layer) and RMS pressure fluctuations, which RANS was incapable of doing. Lastly, a cutting edge DNS simulation added an additional dataset to the community and demonstrated the viability of using highly resolved LES simulations as a surrogate validation dataset. The DNS data also demonstrated the potential of the Bowersox Algebraic Energy Flux turbulence model as a next generation non-Boussinesq model. The presented work demonstrates that while the community has come a very long way, there are still some areas of theoretical improvement required to understand hypersonic wall bounded turbulent flows under the most straining conditions