Search

Stereo vision system and manipulator are two major components of an autonomous fruit harvesting system. In order to raise the fruit-harvesting rate, stereo vision system calibration and kinematic calibration are two significant processes to improve the positional accuracy of the system. This article reviews the mathematics of these two calibration processes and presents an integrated approach for acquiring calibration data and calibrating both components of an autonomous kiwifruit harvesting system. The calibrated harvesting system yields good positional accuracy in the laboratory tests, especially in harvesting individual kiwifruit. However, the performance is not in line with the outcomes in the orchard field tests due to the cluster growing style of kiwifruit. In the orchard test, the calibrations reduce the fruit drop rate but it does not impressively raise the fruit harvesting rate. Most of the fruit in the clusters remain in the canopy due to the invisibility of the stereo vision system. After analyzing the existing stereo vision system, a future visual sensing system research direction for an autonomous fruit harvesting system is justified.

•An explicit moment-based analytical framework coupled with response surface methodology to propagate uncertainties is proposed.•Parameter optimization of polymer electrolyte membrane fuel cell considering uncertainties in inputs.•The framework is accurate and computationally efficient.•High-quality and cost-efficient fuel cell design can be achieved through probabilistic optimization at the expense of optimal power.

Complex number features are ubiquitous in many engineering and scientific applications. Many traditional classification algorithms including alternating decision tree (ADTree) are very powerful but not capable of handling complex domain data. ADTree is classifier that is intrinsically support boosting, hence inherent all desirable statistical properties of boosting methodology. This work introduces base learners that enable application of ADTree algorithm to complex domain data. The presented results show that the proposed base learners enhance performance of ADTrees on complex domain features.

The capability to solve ordinary differential equations (ODE) in hardware will increase the operation capacity of sensing systems in areas such as self-diagnostics, model-based measurement and self-calibration. The computational complexity of solving ODE must be reduced in order to implement a real-time embedded ODE solver. The research proposes a novel design that proves the possibility of solving ODE in real-time embedded systems with reasonably high degree of precision and efficiency. The application of three approximation methods namely, multi-layer perceptron, radial basis network and Lipschitz continuous interpolation is researched and compared.