Optimal VLSI Design for a Highly-Safe Intelligent Vehicle Based on a System Integration Theory
Project/Area Number |
17300009
|
Research Category |
Grant-in-Aid for Scientific Research (B)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
Computer system/Network
|
Research Institution | Tohoku University |
Principal Investigator |
KAMEYAMA Michitaka Tohoku University, Graduate School of Information Sciences, Professor (70124568)
|
Co-Investigator(Kenkyū-buntansha) |
HARIYAMA Masanori Graduate School of Information Sciences, 大学院・情報科学研究科, Associate Professor (10292260)
|
Project Period (FY) |
2005 – 2007
|
Project Status |
Completed (Fiscal Year 2007)
|
Budget Amount *help |
¥13,400,000 (Direct Cost: ¥12,200,000、Indirect Cost: ¥1,200,000)
Fiscal Year 2007: ¥5,200,000 (Direct Cost: ¥4,000,000、Indirect Cost: ¥1,200,000)
Fiscal Year 2006: ¥4,300,000 (Direct Cost: ¥4,300,000)
Fiscal Year 2005: ¥3,900,000 (Direct Cost: ¥3,900,000)
|
Keywords | System-on-Chip / High-level Synthesis / Reconfigurable VLSI / Highly-Safe Intelligent Vehicle / Motion Estimation / Road Extraction / Vehicle Extraction / Human Extraction / ステレオビジョン / メモリアロケーション / 非同期アーキテクチャ / ベイジアンネットワーク / VLSIアーキテクチャ / システムレベル統合設計 / サンプリング周期 |
Research Abstract |
Intelligent algorithms, advanced VLSI architectures and a system integration theory are studied to develop a highly-safe intelligent vehicle. (1) System-Application-Level Design Theory A real-world intelligent integrated system consists of three basic modules of environment recognition, prediction or estimation, and behavior planning. A system integration theory including derivation of performance specification is studied to bridge the gap between the various design levels as well as their VLSI-oriented intelligent algorithms. (2) Processing Module for Highly-Safe Intelligent Vehicles VLSI-oriented algorithms for road extraction, vehicle extraction and human extraction using 3-dimensional image information are developed. Moreover, motion estimation of a vehicle is done based on Bayesian Network The problem is equivalent to estimation of a driver's intention. The driver's intentions are hierarchically defined, so that the designed Bayesian Network becomes as simple as possible. Then, causal relation between the intentions is discussed to reflect the real-world motion process. (3) Optimal Design Theory of VLSI Processors for Intelligent Integrated Systems To achieve power minimization under a time/area constraint, high-level synthesis techniques are investigated based on scheduling and allocation. One typical example is a parallel VLSI for 3-dimensional image processing with optimal memory allocation which solves the data transfer bottleneck between processing elements and memory modules. (4) Reconfigurable VLSI Computing Fine-grained reconfigurable VLSIs for real-world applications superior to the conventional FPGAs are designed and implemented based on new architectures such as direct allocation of a control-data-flow graph, a bit-serial arithmetic operation, a dynamic control of power dissipation, and a logic-in-memory architecture utilizing nonvolatile devices.
|
Report
(4 results)
Research Products
(85 results)