Freeway travel time estimation and prediction using dynamic neural networks

Providing transportation system operators and travelers with accurate travel time information allows them to make more informed decisions, yielding benefits for individual travelers and for the entire transportation system. Most existing advanced traveler information systems (ATIS) and advanced traffic management systems (ATMS) use instantaneous travel time values estimated based on the current measurements, assuming that traffic conditions remain constant in the near future. For more effective applications, it has been proposed that ATIS and ATMS should use travel times predicted for short-term future conditions rather than instantaneous travel times measured or estimated for current conditions. ^ This dissertation research investigates short-term freeway travel time prediction using Dynamic Neural Networks (DNN) based on traffic detector data collected by radar traffic detectors installed along a freeway corridor. DNN comprises a class of neural networks that are particularly suitable for predicting variables like travel time, but has not been adequately investigated for this purpose. Before this investigation, it was necessary to identifying methods for data imputation to account for missing data usually encountered when collecting data using traffic detectors. It was also necessary to identify a method to estimate the travel time on the freeway corridor based on data collected using point traffic detectors. A new travel time estimation method referred to as the Piecewise Constant Acceleration Based (PCAB) method was developed and compared with other methods reported in the literatures. The results show that one of the simple travel time estimation methods (the average speed method) can work as well as the PCAB method, and both of them out-perform other methods. This study also compared the travel time prediction performance of three different DNN topologies with different memory setups. The results show that one DNN topology (the time-delay neural networks) out-performs the other two DNN topologies for the investigated prediction problem. This topology also performs slightly better than the simple multilayer perceptron (MLP) neural network topology that has been used in a number of previous studies for travel time prediction.^

Freeway Travel Time Estimation and Prediction Using Dynamic Neural Networks

Providing transportation system operators and travelers with accurate travel time information allows them to make more informed decisions, yielding benefits for individual travelers and for the entire transportation system. Most existing advanced traveler information systems (ATIS) and advanced traffic management systems (ATMS) use instantaneous travel time values estimated based on the current measurements, assuming that traffic conditions remain constant in the near future. For more effective applications, it has been proposed that ATIS and ATMS should use travel times predicted for short-term future conditions rather than instantaneous travel times measured or estimated for current conditions. This dissertation research investigates short-term freeway travel time prediction using Dynamic Neural Networks (DNN) based on traffic detector data collected by radar traffic detectors installed along a freeway corridor. DNN comprises a class of neural networks that are particularly suitable for predicting variables like travel time, but has not been adequately investigated for this purpose. Before this investigation, it was necessary to identifying methods for data imputation to account for missing data usually encountered when collecting data using traffic detectors. It was also necessary to identify a method to estimate the travel time on the freeway corridor based on data collected using point traffic detectors. A new travel time estimation method referred to as the Piecewise Constant Acceleration Based (PCAB) method was developed and compared with other methods reported in the literatures. The results show that one of the simple travel time estimation methods (the average speed method) can work as well as the PCAB method, and both of them out-perform other methods. This study also compared the travel time prediction performance of three different DNN topologies with different memory setups. The results show that one DNN topology (the time-delay neural networks) out-performs the other two DNN topologies for the investigated prediction problem. This topology also performs slightly better than the simple multilayer perceptron (MLP) neural network topology that has been used in a number of previous studies for travel time prediction.

The impact of solid and liquid wastes from a rural town on the Chorobamba River Oxapampa, Peruvian Amazon

The goal of this study was to determine the instantaneous vs. integrated effects of waste on the water quality of the Chorobamba River. I sampled 9 stations upstream and downstream of the Town of Oxapampa, Peru during the dry season (June-August) of 2004. I measured in-situ parameters such as pH, DO, temperature, etc. as well as vegetation, riverbank erosion, nutrients (N03, NH4, P04), coliform bacteria and macroinvertebrate communities to determine the current conditions of the river, as well as the integrated effects of pollution. Although water quality conditions remained stable, high fecal coliform concentrations and macroinvertebrate communities indicate deterioration in river health over a longer period of time. If riparian areas along the Chorobamba continue to decrease and if inputs of sewage into the rivers continue to increase, as a function of population, then, conditions will continue to deteriorate in the coming years.

Evaluation of contact and non-contact trapping efficiencies of human scent chemical profiles and their stabilities under different environmental conditions

Establishing an association between the scent a perpetrator left at a crime scene to the odor of the suspect of that crime is the basis for the use of human scent identification evidence in a court of law. Law enforcement agencies gather evidence through the collection of scent from the objects that a perpetrator may have handled during the execution of the criminal act. The collected scent evidence is consequently presented to the canines for identification line-up procedures with the apprehended suspects. Presently, canine scent identification is admitted as expert witness testimony, however, the accurate behavior of the dogs and the scent collection methods used are often challenged by the court system. The primary focus of this research project entailed an evaluation of contact and non-contact scent collection techniques with an emphasis on the optimization of collection materials of different fiber chemistries to evaluate the chemical odor profiles obtained using varying environment conditions to provide a better scientific understanding of human scent as a discriminative tool in the identification of suspects. The collection of hand odor from female and male subjects through both contact and non-contact sampling approaches yielded new insights into the types of VOCs collected when different materials are utilized, which had never been instrumentally performed. Furthermore, the collected scent mass was shown to be obtained in the highest amounts for both gender hand odor samples on cotton sorbent materials. Compared to non-contact sampling, the contact sampling methods yielded a higher number of volatiles, an enhancement of up to 3 times, as well as a higher scent mass than non-contact methods by more than an order of magnitude. The evaluation of the STU-100 as a non-contact methodology highlighted strong instrumental drawbacks that need to be targeted for enhanced scientific validation of current field practices. These results demonstrated that an individual's human scent components vary considerably depending on the method used to collect scent from the same body region. This study demonstrated the importance of collection medium selection as well as the collection method employed in providing a reproducible human scent sample that can be used to differentiate individuals.

Dynamic modeling and analysis of single-stage boost inverters under normal and abnormal conditions

Inverters play key roles in connecting sustainable energy (SE) sources to the local loads and the ac grid. Although there has been a rapid expansion in the use of renewable sources in recent years, fundamental research, on the design of inverters that are specialized for use in these systems, is still needed. Recent advances in power electronics have led to proposing new topologies and switching patterns for single-stage power conversion, which are appropriate for SE sources and energy storage devices. The current source inverter (CSI) topology, along with a newly proposed switching pattern, is capable of converting the low dc voltage to the line ac in only one stage. Simple implementation and high reliability, together with the potential advantages of higher efficiency and lower cost, turns the so-called, single-stage boost inverter (SSBI), into a viable competitor to the existing SE-based power conversion technologies.^ The dynamic model is one of the most essential requirements for performance analysis and control design of any engineering system. Thus, in order to have satisfactory operation, it is necessary to derive a dynamic model for the SSBI system. However, because of the switching behavior and nonlinear elements involved, analysis of the SSBI is a complicated task.^ This research applies the state-space averaging technique to the SSBI to develop the state-space-averaged model of the SSBI under stand-alone and grid-connected modes of operation. Then, a small-signal model is derived by means of the perturbation and linearization method. An experimental hardware set-up, including a laboratory-scaled prototype SSBI, is built and the validity of the obtained models is verified through simulation and experiments. Finally, an eigenvalue sensitivity analysis is performed to investigate the stability and dynamic behavior of the SSBI system over a typical range of operation. ^

Beyond the Pay: Current Illicit Activities of the Armed Forces in Central America

The growth of criminal gangs and organized crime groups has created unprecedented challenges in Central America. Homicide rates are among the highest in the world, countries spend on average close to 10 percent of GDP to respond to the challenges of public insecurity, and the security forces are frequently overwhelmed and at times coopted by the criminal groups they are increasingly tasked to counter. With some 90 percent of the 700 metric tons of cocaine trafficked from South America to the United States passing through Central America, the lure of aiding illegal traffickers through provision of arms, intelligence, or simply withholding or delaying the use of force is enormous. These conditions raise the question: to what extent are militaries in Central America compromised by illicit ties to criminal groups? The study focuses on three cases: Nicaragua, El Salvador, and Honduras. It finds that: Although illicit ties between the military and criminal groups have grown in the last decade, militaries in these countries are not yet “lost’ to criminal groups. Supplying criminal groups with light arms from military stocks is typical and on the rise, but still not common. In general the less exposed services, the navies and air forces, are the most reliable and effective ones in their roles in interdiction. Of the three countries in the study, the Honduran military is the most worrying because it is embedded in a context where civilian corruption is extremely common, state institutions are notoriously weak, and the political system remains polarized and lacks the popular legitimacy and political will needed to make necessary reforms. Overall, the armed forces in the three countries remain less compromised than civilian peers, particularly the police. However, in the worsening crime and insecurity context, there is a limited window of opportunity in which to introduce measures targeted toward the military, and such efforts can only succeed if opportunities for corruption in other sectors of the state, in particular in law enforcement and the justice system, are also addressed. Measures targeted toward the military should include: Enhanced material benefits and professional education opportunities that open doors for soldiers in promising legitimate careers once they leave military service. A clear system of rewards and punishments specifically designed to deter collusion with criminal groups. More effective securing of military arsenals. Skills and external oversight leveraged through combined operations, to build cooperation among those sectors of the military that have successful and clean records in countering criminal groups, and to expose weaker forces to effective best practices.

The Effect of Nutrient-Rich Effluents from Shrimp Farming on Mangrove Soil Carbon Storage and Geochemistry Under Semi-Arid Climate Conditions in Northern Brazil

A semi-arid mangrove estuary system in the northeast Brazilian coast (Ceará state) was selected for this study to (i) evaluate the impact of shrimp farm nutrient-rich wastewater effluents on the soil geochemistry and organic carbon (OC) storage and (ii) estimate the total amount of OC stored in mangrove soils (0–40 cm). Wastewater-affected mangrove forests were referred to as WAM and undisturbed areas as Non-WAM. Redox conditions and OC content were statistically correlated (P < 0.05) with seasonality and type of land use (WAM vs. Non-WAM). Eh values were from anoxic to oxic conditions in the wet season (from − 5 to 68 mV in WAM and from < 40 to > 400 mV in Non-WAM soils) and significantly higher (from 66 to 411 mV) in the dry season (P < 0.01). OC contents (0–40 cm soil depth) were significantly higher (P < 0.01) in the wet season than the dry season, and higher in Non-WAM soils than in WAM soils (values of 8.1 and 6.7 kg m− 2 in the wet and dry seasons, respectively, for Non-WAM, and values of 3.8 and 2.9 kg m− 2 in the wet and dry seasons, respectively, for WAM soils; P < 0.01). Iron partitioning was significantly dependent (P < 0.05) on type of land use, with a smaller degree of pyritization and lower Fe-pyrite presence in WAM soils compared to Non-WAM soils. Basal respiration of soil sediments was significantly influenced (P < 0.01) by type of land use with highest CO2 flux rates measured in the WAM soils (mean values of 0.20 mg CO2 h− 1–g− 1 C vs. 0.04 mg CO2 h− 1–g− 1 C). The OC storage reduction in WAM soils was potentially caused (i) by an increase in microbial activity induced by loading of nutrient-rich effluents and (ii) by an increase of strong electron acceptors [e.g., NO3−] that promote a decrease in pyrite concentration and hence a reduction in soil OC burial. The current estimated OC stored in mangrove soils (0–40 cm) in the state of Ceará is approximately 1 million t.

Tree Islands in Everglades Landscapes: Current Status, Historical Changes, and Hydrologic Impacts on Population Dynamics and Moisture Relations, First Annual Report

In 2005 we initiated a project designed to better understand tree island structure and function in the Everglades and the wetlands bordering it. Focus was on the raised portions at the upstream end of the islands, where tropical hardwood species adapted to well-drained conditions usually are the most prominent component of the vegetation. The study design is hierarchical, with four levels; in general, a large number of sites is to be surveyed once for a limited set of parameters, and increasingly small sets of islands are to be sampled more intensively, more frequently, and for more aspects of ecosystem function. During the first year of the 3-year study, we completed surveys of 41 Level 1 (i.e., the least intensive level) islands, and established permanent plots in two and three islands of Levels 2 and 4 intensity, respectively. Tree species richness and structural complexity was highest in Shark Slough “hammocks”, while islands in Northeast Shark Slough and Water Conservation Area 3B, which receive heavy human use, were simpler, more park-like communities. Initial monitoring of soil moisture in Level 4 hammocks indicated considerable local variation, presumably associated with antecedent rainfall and current water levels in the adjacent marsh. Tree islands throughout the study area were impacted significantly by Hurricanes Katrina and Wilma in 2005, but appear to be recovering rapidly. As the project continues to include more islands and repeated measurements, we expect to develop a better grasp of tree island dynamics across the Everglades ecosystem, especially with respect to moisture relations and water levels in the adjacent marsh. The detailed progress report which follows is also available online at http://www.fiu.edu/~serp1/projects/treeislands/tree_islands_2005_annual_report.pd