In two experiments pigeons were exposed to VI and FI schedules and viii schedules approximating both VI and FI schedules. In experiment I, the probabilities of the VI and FI components in a Mixed FI VI schedule were manipulated to create schedule contingencies approximating simple VI or FI. In experiment II, the minimum and maximum inter-reinforcement intervals were manipulated to create schedule contingencies approximating simple VI or FI. The major finding of both of these experiments was that maximal control by the dimension of time occurred as FI contingencies were approximated. Control by any one temporal value in experiment I depended on its temporal separation from 100 seconds and the probability of reinforcement associated with 100 seconds. Control by any one temporal value in experiment II depended on its temporal separation from the minimum inter-reinforcement interval and 100 seconds. The results were discussed in terms of interval schedule control as a form of stimulus control. A third experiment was performed to examine possible relationships between the pause in FI performance and the subsequent scallop. The baseline condition was contaminated by a procedure which may have produced effects which overrode experimental manipulations. Finally, an experiment was suggested to demonstrate behavioral contrast along a temporal dimension. The argument that interval schedule control is a form of stimulus control rested on analogy and inference. A demonstration of behavioral contrast along a temporal dimension would demonstrate more directly that time is similar to other dimensions. Hence, the same principles could be used to explain schedule control as are used to explain stimulus control.

The present study was designed to investigate the effects of a stimulus change inserted in the large component of a mixed fixed-ratio 10 fixed-ratio 100 schedule. In mixed fixed-ratio schedules, a run of responses which approximates the response requirement of the smaller fixed ratio occurs at the beginning of the larger fixed ratio. This run of responses and the pause it precedes is called a prime. In Experiment I, priming acquisition was compared in a mixed schedule in which a change in key-color followed completion of the first 10 responses of the fixed-ratio 100 component and a mixed schedule with no stimulus change. Primes were acquired more rapidly in the mixed schedule with the stimulus change and the number of responses in a priming run was less variable than occurred without the stimulus change. In Experiment II, the effect of the stimulus change on primes was further investigated by removal of the smaller fixed-ratio component or by varying the location of the stimulus change. Primes occurred only when the smaller fixed-ratio component was present. Varying the location of the stimulus change resulted in the transfer of control from the external stimulus to response-produced stimuli. Future investigation of this point of transfer should prove useful in the study of the proprioceptive stimulus control of homogeneous behavior sequences. In Experiments Ill through VI, the variables controlling the length of the post-prime and the post-reinforcement pauses were investigated using the mixed FR x chained FR x FR y schedule. In this series of experiments, FR x was varied with FR y held constant at both high and low values. In addition, FR y was varied with FR x held constant at both high and low values. The results indicate that the post-prime pause is primarily a function of FR y, the number of responses required after the priming run. On the other hand, both post-reinforcement pauses were shown to be a function of FR x, the size of the small fixed ratio. An interaction between FR x and FR y and both the post-reinforcement and the post-prime pauses suggested that pausing in mixed schedules is a closed system. Taken as a whole, these results indicate the importance of mixed chained schedules in the investigation of the priming phenomenon.

This research examines the drying of hardwood timber with particular reference to seasoning blackbutt (Eucalyptus pilularis) in a solar kiln. The aims of this research were to develop an optimised drying schedule for drying blackbutt and to develop and validate a mathematical model for a solar kiln. In the first stage of this study, the cross-grain physical and mechanical properties were determined for blackbutt timber so that an optimised schedule (based on drying within a limiting strain envelope) can be developed using model predictive control techniques for drying 43 mm thick (green) blackbutt timber boards in solar kilns. This optimised schedule has been developed and tested in the laboratory. The drying time was 10% shorter for this schedule than the original schedule, compared with an expected reduction in drying time of 14% (relative to the original schedule). Overall the quality was slightly better and the drying time was shorter for the optimised schedule compared with the original schedule. A complete solar kiln model has been developed and validated based on comparisons between the predicted and the measured internal air temperatures, relative humidities and timber moisture contents. The maximum difference between the actual and predicted moisture contents was 0.05 kg kg-1. The agreement between the predicted and measured temperatures of the internal air is reasonable, and both the predictions and measurements have a similar cyclical pattern. The generally good agreement between the model prediction of the final moisture content and its measurement may be due to the careful measurement of the boundary conditions such as the solar energy input. The key uncertainties were identified as the heat exchanger output, the measurement of the initial moisture content, the amount of accumulation of condensate on the floor, and the estimation of sky temperature. The significant uncertainty (18%) in the estimation of the initial moisture content is a key reason for the mismatch between the model prediction and the measurements. In terms of operating variables, the energy release rate from the heat exchanger had the greatest effect on the simulated performance, followed by the water spray and venting rates. The simulation suggested that a material with a lower transmissivity to thermal radiation may effectively lower radiation losses, improving the kiln performance, so such materials for glazing is a high priority.

The literature reviewed on lot-sizing models with random yields is limited to certain random occurrences such as day to day administrative errors, minor machine repairs and random supply due to faulty delivery of parts. In reality however, the manufacturing industry faces other risks that are non random in nature. One example would be yield discrepancies caused by non random triggers such as a change in the production process, product or material. Yield uncertainties of these types are temporary in nature and usually pertain until the system stabilizes. One way of reducing the implications of such events is to have additional batches processed earlier in the production that can absorb the risk associated with the event. In this thesis, this particular approach is referred to as the <i>anticipatory batch insertion</i> to mitigate perceived risk. This thesis presents an exploratory study to analyze the performance of batch insertion under various scenarios. The scenarios are determined by sensitivity of products, schedule characteristics and magnitude of risks associated with causal triggers such as a process change. The results indicate that the highest return from batch insertion can be expected when there are slightly loose production schedules, high volumes of sensitive products are produced, there are high costs associated with the risks, and the risks can be predicted with some degree of certainty.

The implementation of automatic transit data collection via Automatic Vehicle Location (AVL) and Automatic Passenger Counting (APC) systems provides an opportunity to create large, detailed datasets of transit operations. These datasets are valuable because they provide an opportunity to evaluate and optimize transit operations using methods that were previously infeasible and without the need for expensive manual data collection. This thesis develops a methodology to utilize data collected by typical AVL/APC system installations in order to (a) develop advanced performance measures to quantify schedule adherence and (b) automatically determine the causes of poor schedule adherence. The methodology addresses the difficulty that many small to medium sized transit agencies have in utilizing the data being collected by proposing a methodology that can be automated, thereby reducing resource and expertise requirements and allowing the data to be more effectively utilized. The ultimate output of the proposed methodology includes the following: 1. A ranked list of routes by direction (for a given time period) that identifies routes with the poorest schedule adherence performance. 2. Performance measures within any given route, direction, and time period that identify which timepoints are contributing most to poor schedule adherence. 3. Statistics indicating identified causes of poor schedule adherence at individual timepoints. 4. A visualization aid to be used in conjunction with the cause statistics generated in Step 3 in order to develop an effective strategy for improving schedule adherence issues. With this information, transit agencies will be able to act proactively to improve their transit system, rather than wait until they discover problems on their own or hear complaints from passengers and drivers. The methodology is tested and demonstrated through application to AVL/APC system data from Grand River Transit, a public transit agency serving Waterloo Region in Ontario, Canada.

The literature reviewed on lot-sizing models with random yields is limited to certain random occurrences such as day to day administrative errors, minor machine repairs and random supply due to faulty delivery of parts. In reality however, the manufacturing industry faces other risks that are non random in nature. One example would be yield discrepancies caused by non random triggers such as a change in the production process, product or material. Yield uncertainties of these types are temporary in nature and usually pertain until the system stabilizes. One way of reducing the implications of such events is to have additional batches processed earlier in the production that can absorb the risk associated with the event. In this thesis, this particular approach is referred to as the <i>anticipatory batch insertion</i> to mitigate perceived risk. This thesis presents an exploratory study to analyze the performance of batch insertion under various scenarios. The scenarios are determined by sensitivity of products, schedule characteristics and magnitude of risks associated with causal triggers such as a process change. The results indicate that the highest return from batch insertion can be expected when there are slightly loose production schedules, high volumes of sensitive products are produced, there are high costs associated with the risks, and the risks can be predicted with some degree of certainty.

The implementation of automatic transit data collection via Automatic Vehicle Location (AVL) and Automatic Passenger Counting (APC) systems provides an opportunity to create large, detailed datasets of transit operations. These datasets are valuable because they provide an opportunity to evaluate and optimize transit operations using methods that were previously infeasible and without the need for expensive manual data collection. This thesis develops a methodology to utilize data collected by typical AVL/APC system installations in order to (a) develop advanced performance measures to quantify schedule adherence and (b) automatically determine the causes of poor schedule adherence. The methodology addresses the difficulty that many small to medium sized transit agencies have in utilizing the data being collected by proposing a methodology that can be automated, thereby reducing resource and expertise requirements and allowing the data to be more effectively utilized. The ultimate output of the proposed methodology includes the following: 1. A ranked list of routes by direction (for a given time period) that identifies routes with the poorest schedule adherence performance. 2. Performance measures within any given route, direction, and time period that identify which timepoints are contributing most to poor schedule adherence. 3. Statistics indicating identified causes of poor schedule adherence at individual timepoints. 4. A visualization aid to be used in conjunction with the cause statistics generated in Step 3 in order to develop an effective strategy for improving schedule adherence issues. With this information, transit agencies will be able to act proactively to improve their transit system, rather than wait until they discover problems on their own or hear complaints from passengers and drivers. The methodology is tested and demonstrated through application to AVL/APC system data from Grand River Transit, a public transit agency serving Waterloo Region in Ontario, Canada.

In order to increase the data-path utilization of the programmable graphics processor units (GPU) which often stall by waiting for the execution results of those long-latency instructions, multi-thread technique is very often used in the design of GPU. This thesis proposes a multi-thread single unified core GPU design which owns several key features. First, its processor core can execute not only the vertex and fragment shading programs, but also the software rasteriation module which is mostly implemented by a individual hardware module in other GPU designs. Next, the thread-switching policy in our design is based on the non-preempt blocked scheduling. Normally, whether an instruction will be stalled cannot be detected until it enters the instruction-decode stage. In order to achieve zero-penalty thread switching, a single assistant bit will be padded to each instruction in a thread to tell if the next instruction in the same thread will be stalled or not. This mechanism can help achieve a speed-up of 1.4 in some benchmarks used in this thesis. The register file used in GPU processor is usually equipped with up to four access ports, such that it will occupy a significant portion of the entire GPU especially for muti-thread designs where the register set has to be duplicated by several copies. The implementation cost of the register file can be reduced by decreasing its access port number to two based on the proposed multi-bank approach in this thesis. Our experimental results show that this approach can help reduce the overall gate count by 26.12%. Finally, the rest of fixed-pipeline fragment operation is realized by an iterative time-sharing architecture in order to further save the silicon area. The overall gate count of the proposed GPU is 600K.

During the past two decades, it was witness a dramatic global shift in economic policy away from state-owned enterprises (SOEs) towards privatization. Since privatization improves incentives, a rapid transfer of ownership and control right should be desirable. In Taiwan, the Executive Yuan of the government organized a group to promote privatization in 1989 and the officers then took the initiative to carry out the privatization program enthusiastically. By classifying the privatization process into two stages, we analyzed the effect of different stock floatation schedules, different underwriting mechanisms, and different levels of government intervention on ownership structure and corporate governance. Based on the model inference, we found that when maximizing stock floating revenue, wider share ownership, and promoting the SOEs¡¦ efficiency, a sequential transfer of ownership and control right should be better than privatized instantaneously. Under the same goals, the mechanism with partial public offering and partial auction was better than the mechanism with partial public offering and partial book building. Finally, it was not optimal for government to intervene the operating of SOEs after privatization.

With technology increasingly and the needs of high performance and multiple functionalities, power dissipation has be a bottleneck in microprocessors. And clock power is the most percentage of total power dissipation. In our thesis, we will provide an effective clock gating methodology that has not more overhead possibly to decrease total power dissipations based on SIMD 3D vertex shader. Except for classify all instructions according the instruction flow, we also consider the relationship of pipeline stage and are based on register bank to control execution units more flexibility. Using clock gating not only can decrease clock power, but also decrease the power of hardware modules succeed the registers with clock gating that be controlled. In our thesis, we will analysis which clock gating version is suitable because there is not definitely to disable the clock of all pipeline registers of all pipeline stages and hold all opportunities that can disable the clock. We will explain on experimental results and show the final low power version. With experimental results, the clock gating methodology that we bring can decrease almost 30% power with increase less than 2% area. And collection of instruction schedule algorithm for high performance that can decrease 41% energy at most. In new version of four vertexes execute sequentially, using clock gating can also decrease almost 10% power dissipation. And collection of instruction schedule algorithm for this version not only has better performance result but also can decrease 16% energy at most.