<p>Diesel engines emit toxic gases like NOx and hydrocarbons. These gases need to be treated before they are released out the tailpipe. Thus, an aftertreatment system is installed which comprises of DOC, DPF and SCR. The DOC oxidizes the hydrocarbons and NO, the DPF traps the particulate matter and SCR reduces the NOx by reacting with NH3 at high temperatures. However, since NH3 is also a toxic gas, it cannot be released out the tailpipe in excess. It is important to inject an appropriate amount of NH3 so that it does not slip out the tailpipe. With increasingly stringent regulations on the emission limits of these toxic gases, control of SCR has become more necessary than before.</p> <p>In this thesis, the work done by previous members of the lab research group was improved upon. The objective remained the same, namely, keeping the NH3 slip under 50 ppm while maximizing NOx reduction. On initial inspection, it was realized that the entire controller had been designed and implemented in continuous time. Since the controller would be implemented digitally, with limited hardware sampling time, a discrete-time implementation as done via a DCU was created. The controller switched between two controllers – slip-based and storage-based. The slip-based controller was modified to include a feedforward term in the system so that the response time could be improved along with a feedback controller to eliminate any disturbances and steady-state error, using ammonia slip feedback as measured by an NH3 sensor. It aims at keeping the maximum ammonia slip under 50 ppm. The storage controller is a feedback controller which tries to limit the ammonia storage based on the values fed by a lookup table. This lookup table is a simplified table that determines the maximum ammonia storage at any given instant based on the catalyst bed temperature. The feedback controller gains for both controllers were determined based on a linearized plant model since the initial gains were ineffective with the discretized model. The initial switching mode controller that switches between slip control and storage control switched too frequently between the controllers, thereby affecting controller performance. A switching logic was implemented to limit the number of switches. A switch will be permitted only if the previous switch occurred over a certain time. By implementing all the subparts together in the controller, incremental improvements were prominent. In the end, the performance by implementing the proposed idea was distinctly better. The metrics considered for performance comparison are the number of switches and the ability to maximize slip up to 50 ppm. Parameter error was also studied as well and its effect on the controller performance was analyzed. The data when tested against sets of underestimated, overestimated and mixed estimates for the plant parameters resulted in the underestimated parameters to work within the scope of the objective. The controller was able to compensate for the underdosing. Overestimation caused overdosing in the system which led to spikes in the NH3 slip. Thus, it is better to underestimate the plant parameters than overestimate them.</p>

The aim of this thesis project is to explore a network traffic regulator using bandwidth management techniques that regulates data traffic with the objective to use the network bandwidth to its maximum capacity while ensuring that the network is not overloaded. The bandwidth in the existing network architecture is shared between two co-existing, distinct data flows for on-board communication and diagnostic communication in an in-vehicle network. The diagnostic communication must not interfere with the more critical on-board communication and it should comply with the remaining bandwidth. In the existing solution, fixed delays are imposed on the data traffic which result in a waste of network capacity. The approach presented in this thesis uses two regulation algorithms for different types of diagnostic services. One regulation algorithm is activated for diagnostic services that require data segmentation and multiple data frames to accommodate the transferred data. This algorithm makes use of the Flow Control parameter Separation Time specified in ISO 15765-1:2011 "Road vehicles -- Diagnostic communication over Controller Area Network (DoCAN)". The other algorithm regulates diagnostic services that generate bursts of single frames where data segmentation is not required and it does so using traffic shaping techniques. The results in this thesis show that the network traffic indeed can be regulated for different diagnostic services by using the two mentioned regulation algorithms. The results also show that data is not lost due to high network utilisation and that the bandwidth is used to its maximum capacity without having to impose fixed delays on the network system. The regulator is adaptive in the sense that it can be used for different vehicle configurations with compatible network systems to ensure quality of service and a robust network system. / I detta examensarbete är målet att utforska en metod för att reglera  nätverkstrafik genom att använda tekniker inom bandbreddshantering  med syfte att utnyttja bandbredden till dess maximala kapacitet utan att överbelasta nätverket. Bandbredden i den nuvarande nätverksarkitekturen delas mellan två dataflöden för onboard kommunikation och diagnostisk kommunikation. Den diagnostiska kommunikationen får inte på någotvis störa onboard kommunkationen och får anpassa sig till den bandbredd som kvarstår. I det existerande systemet införs fixa fördröjningar i nätverkstrafiken vilket medför ett onödigt slöseri på nätverkskapaciteten och som också medför att de diagnostiska tjänsterna tar längre tid att utföra.  Tillvägagångssättet som presenteras i detta arbete använder två regleringsalgoritmer för olika typer av diagnostiska tjänster. En algoritm används för tjänster som kräver datasegmentering och flera dataramar för att skicka data. Den här algoritmen använder parametern Separation Time som är specificerad i ISO standarden 15765-1:2011 "Road vehicles -- Diagnostic communication over Controller Area Network (DoCAN)". Diagnostiska tjänster som istället genererar en skur av enstaka dataramar regleras med en traffic shaping algoritm som heter Token Bucket. Resultaten i detta arbete visar att det går att reglera nätverkstrafiken för olika typer av diagnostiska tjänster genom att använda de två utvecklade algoritmerna. Resultaten visar också att data inte går förlorat vid höga nätverkslaster och att bandbredden används maximalt utan att behöva införa fixa fördröjningar i nätverkssystemet. Regleraren är adaptiv i bemärkelsen att den kan användas för alla tänkbara fordonskonfigurationer med kompatibelt nätverkssystem för att försäkra quality of service och robusthet.

The Atlanta Inspection and Maintenance program ultimately aims to reduce on-road vehicular emission, a major source of air pollution. The program enforces eligible vehicles to be inspected and repaired, if necessary, before the annual registration renewal. However, various factors can influence the program implementation with respect to the motorists, inspectors, and testing technology. This research explores some of these factors by using empirical data from the Continuous Atlanta Fleet Evaluation project, the inspection transaction records, the Atlanta Household Travel Survey, and the U.S. Census Bureau. The study discusses policy implications of findings from the three essays and offers related recommendations. The first essay examines whether the higher income of a vehicle owner decreases the odds of the vehicle failing the first inspection. Findings show that vehicles owned by low-income households are more likely to fail the first inspection of the annual test cycle. However, after controlling for the vehicle characteristics, the odds of failing the first inspection are similar across households. This suggests that the maintenance behaviors are approximately the same for high- and low-income households. The second essay explains the motorists' decisions in selecting their inspection stations using a random utility model. The study finds that motorists are likely to choose the inspection stations that are located near their houses, charge lower fees, and can serve a large number of customers. Motorists are less likely to choose the stations with a relatively high failure ratio especially in an area of low station density. Moreover, motorists do not travel an extra mile to the stations with lower failure ratio. Understanding choices of vehicle owners can shed some light on the performance of inspection stations. The third essay investigates the validity and reliability of the on-board diagnostic generation II (OBD II) test, a new testing technology required for 1966 and newer model year vehicles. The study compares the inspection results with the observed on-road emission using the remote sensing device (RSD) of the same vehicles. This research finds that the agreement between the RSD measurement and the OBD II test is lower for the relatively older or higher use vehicle fleets