81 research outputs found
Chance-Constrained Control with Imperfect Perception Modules
No full textAutonomous systems are required to operate in different environments, but recognizing the current environment is often challenging. For example, an autonomous vehicle should stop or obey a speed limit according to a traffic sign, but state-of-the-art perception modules (e.g., neural networks) do not guarantee the correctness of their reading of the traffic sign. Considering such uncertain outputs of a perception module, which in effect determines modes, we propose a chance-constrained control formulation that with high probability guarantees the satisfaction of a set of constraints associated with the possible modes. To do this, we present a method based on the Bayes rule and sampling to calculate the probability of each mode. We prove that our approach can ensure satisfying constraints of novel situations, which have not been used during training of the perception module. Also, to account for the error due to limited data, we present a robust formulation that guarantees constraint satisfaction with high confidence. In an autonomous vehicle example, we train a neural network that classifies traffic signs and show that given each output of the neural network, our motion planning approach guarantees the constraint satisfaction with high probability
Diffusion of Biological Organisms: Fickian and Fokker-Planck Type Diffusions
No full textIn this paper we derive diffiusion equations in a heterogeneous environment. We consider a system of discrete kinetic equations that consists of two phenotypes of different turning frequencies. The two phenotypes change their states according to state transition frequencies which depend on the environment. We show that the density of the total population of the two phenotypes converges to the solution of a Fokker-Planck type diffiusion equation if turning frequencies are of higher order than the state transition frequencies. If it is the other way around, i.e., if the state changes many times between each turning, the density converges to the solution of a Fickian diffiusion equation.11Nsciescopu
Adaptive Thermoelectric Cooling Arrays for Energy-Efficient Local and Transient Heat Management
No full text1
Automated Driving System with Guaranteed Safety based on Generic Environment Representation and Model Predictive Control
No full text학위논문 (박사)-- 서울대학교 대학원 : 기계항공공학부, 2016. 2. 이경수.Recently, the interest of automotive researches changes from the passive safety system to the active safety system and, by extension, automated driving system due to advances in sensing technologies. For example, active safety applications, such as vehicle stability control (VSC), adaptive cruise control (ACC), lane keeping assistance (LKA) and lane change assistance (LCA) system), automated parking assist system (APA) and blind spot intervention (BSI), already have been commercialized by major automakers Furthermore, there are various ongoing projects which are trying to achieve the zero fatality. Several research teams around the world are continuously advancing the field of autonomous driving. And some of major automakers have been researching to integrate individual active safety system for the enhancement of safety. GM is trying to develop and introduce Super Cruise system which can drive on the highway without human drivers intervention. Toyota has undertaken researches to develop Automatic Highway Driving Assist technology. The BMW managed to drive 100% automated in real traffic on the freeway from Munich to Ingolstadt, showing a robust, comfortable, and safe driving behavior, even during multiple automated LC maneuvers and the Mercedes Benz developed Intelligent Drive system and followed the route from Mannheim to Pforzheim, Germany, in fully autonomous manner
From a careful review of considerable amount of literature, automated driving technology has the potential to reduce the environmental impact of driving, reduce traffic jams, and increase the safety of motor vehicle travel. However, the current state-of-the-art in automated vehicle technology requires precise, expensive sensors such as differential global positioning systems, and highly accurate inertial navigation systems and scanning laser rangefinders. While the cost of these sensors is going down, integrating them into cars will increase the price and represent yet another barrier to adoption.
Therefore, this dissertation focused on developing a fully automated driving algorithm which is capable of automated driving in complex scenarios while a chosen sensor configuration is closer to current automotive serial production in terms of cost and technical maturity than in many autonomous vehicles presented earlier. Mainly three research issues are considered: an environment representation, a motion planning, and a vehicle control.
In the remainder of this paper, we will provide an overview of the overall architecture of the proposed automated driving control algorithm and the experimental results which shown the effectiveness of the proposed automated driving algorithm. The effectiveness of the proposed automated driving algorithm is evaluated via vehicle tests. Test results show the robust performance on an inner-city street scenario.Chapter 1 Introduction 1
1.1. Background and Motivation 1
1.2. Previous Researches 4
1.3. Thesis Objectives 7
1.4. Thesis Outline 8
Chapter 2 Overview of an Automated Driving System 9
Chapter 3 Environment Representation 12
3.1. Driving Corridor Decision 14
3.2. Static Obstacle Map Construction 19
Chapter 4 Moving Object Tracking and Estimation 21
4.1. Problem Formulation 22
4.1.1. Stochastic hybrid system 22
4.1.2. Coordinate Systems 24
4.1.3. Standard Process Model 25
4.1.4. Standard Measurement Model 28
4.2. Selection of Multiple Model Set and Parameter Design 31
4.2.1. Set of Multiple Process Model 31
4.2.2. Set of Multiple Measurement Model 33
4.2.3. Event Dependent Transition Probability Matrix 35
4.3. IMM/EKF Multi Target State estimation 40
4.3.1. Host Vehicle Filter 41
4.3.2. IMM/EKF based Filtering 42
4.3.3. Track Management 45
4.4. Vehicle Tests based Performance Evaluation 47
4.4.1. Configuration of Vehicle Tests 47
4.4.2. Implementation and Evaluation 49
4.4.3. Comparison with Model-switching/EKF 54
4.4.4. Experimental Results with Multi-target Situation 57
Chapter 5 . Safety Driving Envelope Decision and Motion Optimization 63
5.1. Multi-traffic Prediction 64
5.1.1. Lane Keeping Behavior Model 66
5.1.2. Vehicle Predictor 68
5.1.3. Test Data based Implementation and Performance Evaluation 72
5.2. Safety Driving Envelope Decision 83
5.3. Model Predictive Control based Motion Planning 86
Chapter 6 Vehicle Tests based Performance Evaluation 90
6.1. Test-Data based Simulation 91
6.2. Vehicle Tests: Automated Driving on Urban Roads 98
Chapter 7 Conclusions 106
Bibliography 108
Abstract in Korean 114Docto
Distinct roles for energy storage and transmission infrastructure in a renewables-based electric power system
No full textDue to the intermittency of renewable resources, achieving a high coverage of renewable generation at low cost is one of the main hurdles to realizing zero-carbon electricity generation. In this study, we analyze the roles of energy storage systems (ESS) and transmission infrastructure in the cost-optimal deployment of a renewable electricity grid in the United States. We find that storage and transmission serve distinctly different functions: transmission is useful for addressing hours-long resource lows, but only plays a supplementary role in mitigating long-duration resource lows. Conversely, storage can handle both short-duration and long-duration resource lows. These different functions are driven in part by the large spatial footprints of the most extreme long duration resource lows. Furthermore, the total cost of renewable energy in the system and the cost-determining technological components in the system are dependent on the renewables penetration toward total demand—known as the energy availability factor (EAF). When the EAF is sufficiently low, the cost of a cost-optimized system is driven solely by generation costs. For low to intermediate EAF, both generation and transmission costs are dominant factors. At high EAF, generation and storage costs become the dominant factors.S.M
LazyRS: Improving the Performance and Reliability of High-Capacity TLC/QLC Flash-Based Storage Systems Using Lazy Reprogramming
No full textWe propose a new NAND programming scheme called the lazy reprogramming scheme (LazyRS) which divides a program operation into two stages, where the second stage is delayed until it is needed. LazyRS optimizes the program latency by skipping the second stage if it is not required. An idle interval before the second stage improves the flash reliability as well. To maximize the benefit of LazyRS, a LazyRS-aware FTL adjusts the length of an idle interval dynamically over changing workload characteristics. The experimental results show that the LazyRS-aware FTL can efficiently improve the write throughput and reliability of flash-based storage systems by up to 2.6 times and 31.2%, respectively
REO: Revisiting Erase Operation for Improving Lifetime and Performance of Modern NAND Flash-Based SSDs
No full textThis work investigates a new erase scheme in NAND flash memory to improve the lifetime and performance of modern solid-state drives (SSDs). In NAND flash memory, an erase operation applies a high voltage (e.g., >20 V) to flash cells for a long time (e.g., >3.5 ms), which degrades cell endurance and potentially delays user I/O requests. While a large body of prior work has proposed various techniques to mitigate the negative impact of erase operations, no work has yet investigated how erase latency and voltage should be set to fully exploit the potential of NAND flash memory; most existing techniques use a fixed latency and voltage for every erase operation, which is set to cover the worst-case operating conditions. To address this, we propose Revisiting Erase Operation, (REO) a new erase scheme that dynamically adjusts erase latency and voltage depending on the cells’ current erase characteristics. We design REO by two key apporaches. First, REO accurately predicts such near-optimal erase latency based on the number of fail bits during an erase operation. To maximize its benefits, REO aggressively yet safely reduces erase latency by leveraging a large reliability margin present in modern SSDs. Second, REO applies near-optimal erase voltage to each WL based on its unique erase characteristics. We demonstrate the feasibility and reliability of REO using 160 real 3D NAND flash chips, showing that it enhances SSD lifetime over the conventional erase scheme by 43% without change to existing NAND flash chips. Our system-level evaluation using eleven real-world workloads shows that an REO-enabled SSD reduces average I/O performance and read tail latency by 12% and 38%, respectivley, on average over a state-of-the-art technique
Development and Application of Team Communication Program and it’s Effects on Sport Weness and Achievement
No full text- …
