| Project/Area Number |
60860025
|
|---|
| Research Category |
Grant-in-Aid for Developmental Scientific Research
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
農業機械
|
|---|
| Research Institution | Kyoto University |
|---|
Principal Investigator |
IKEDA Yoshio Kyoto University, Faculty of Agriculture, 農学部, 助教授 (50026465)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
GOTOH Kiyokazu Kyoto University, Faculty of Agriculture, 農学部, 助手 (60026581)
|
|---|
| Project Period (FY) |
1985 – 1987
|
| Project Status |
Completed (Fiscal Year 1987)
|
| Budget Amount *help |
¥11,600,000 (Direct Cost: ¥11,600,000)
Fiscal Year 1987: ¥1,500,000 (Direct Cost: ¥1,500,000)
Fiscal Year 1986: ¥600,000 (Direct Cost: ¥600,000)
Fiscal Year 1985: ¥9,500,000 (Direct Cost: ¥9,500,000)
|
|---|
| Keywords | Quality og Agricultural Products / Nondestructive Inspection / Linear Prediction Model Maximum Entropy Method / Kalman Filter / Discriminant Apalysis / Shape Memory Alloy / 形状記憶合金 / メカニカルグリッパー / 周波数応答関数 / 線形システム / ARMAモデル / ARモデル / 平均応答法 / 農業用ロボット / 人工触覚 / ARMA法 / 不規則加振 / 農業用ロボットハンド |
|---|
| Research Abstract |
For the Uanatitative evaluation of Quality change of the agricultural products during storage with their dynamic response, the experimental tandom vibrator was produced for the whole intact products. This apparatus was computer controlled to aquire and process the data automatically. The linear prediction models were adopred to describe the dynamic behavior of the products, that is, the autoregressive model(AR(10) mode) or the autoregressive-moving average mode(ARMA(4,3) model), and the parametric representations of the dynamics of the products were proposed. The parameters of the AR and ARMA model were estimated with the Bur'g Maximum Entropy Method(MEM) and the Recursive Least Square Method(RLSM) based upon the Kalman Filtering, respectively. It was shown that the ARMA model containing the smaller number of terms than the AR model could simulate the frequency response functions of the products estimated with the FFT method.
Though the prediction coefficients of the AR or ARMA model ma
… More
y be assumed to express quantitatively the dynamic characteristics of the products, it may be difficult to relate the product quality to these parameters intutively because of their multi-demensionality. Then, the discriminant analysis of the quality change during storage was performed with the Mahalanobis's generalized distance. The results of the discriminant analysis were fairly good it seemed that the discriminant was performed under the influences of the variance of the paramters between samples, The improvement of the experimental appsratus and the realization of the stable excitation will be necessary.
It was shown that the change in this mahalanobis distance during storage could be used as quantitative indexes representing the quality change. This distance inctreased during storage and their absolute values became longer under the unfavorable storage conditions.
The experimental gripper equiped the driving device made of the shape memory alloy was produced for handling the agricultural products. It was shown through the fundamental experiment that this gripper had the enough grasping force but the slow response speed. Less
|
