Suming Chen

RFID-integrated Multi-Functional Remote Sensing System for Seedling Production Management

A multi-functional remote sensing system (MFRSS) integrating Radio Frequency Identification (RFID) technology with remote spectral imaging and environmental sensing was developed to enhance seedling production and management in greenhouse. Consisted of a management traceability system (also known as management resume) and an environment traceability system (also known as environment resume), a traceable production management system is highly desirable for automated operations in greenhouse. This study is the first application of RFID to an automatic greenhouse seedling production system. The advantages of the MFRSS system are twofold: First, the production management information can be made traceable by establishing passive RFID on the multi-functional remote sensing system during variable boom cruising in the greenhouse. The spectral images were acquired using a color camera and a monochrome camera with a 780 nm optical filter, with the exposure time and signal gain controlled through an IEEE-1394 interface. Secondly, an automatic exposure algorithm eliminating interferences caused by sunlight variation was developed using Matlab 6.5 and LabVIEW 7.1. The tray position data were transferred to the look up table and delivered to the water management module through a DataSocket server and wireless network. The environment-sensing sub-system, including temperature, relative humidity, and photo-quantum measurements, was developed with a PCI-6023 interface to analyze their spatial distribution in the greenhouse. Not only does the mechanism established in the present study provide a basis for developing an automatic seedling production system, the environmental factors and facility status captured by the RFID-integrated MFRSS also enable a traceable seedling production management database, which is a crucial constituent for practical applications.

Development of a tubing-grafting robotic system for fruit-bearing vegetable seedlings.

This research aims to develop an automatic tubing-grafting robotic system using soft rubber tubes as the grafting material that is suitable for the rootstock and scion of fruit-bearing vegetables having similar stem diameter, such as tomato, bitter melon and sweet pepper. Compared with grafting using grafting clips, the soft rubber tube approach is low-cost and minimizes water loss from the grafting wounds. Moreover, as the grafted seedlings grow, the soft rubber tubes automatically peel off from seedlings due to embrittlement and this saves recycling. The tubing-grafting robotic system is made up of six major units, namely the rootstock gripping, the rootstock cutting, the scion gripping, the scion cutting, the tube supplying, and the guiding units. A pneumatic mechanism with a programmable logic controller is used to process the sequential events needed. In the guiding unit, a pair of guiding sleeves ensures that the rootstock and scion are able to be inserted into the rubber tube smoothly even if the stem of either the rootstock or scion is deflective. The experimental results indicated that the average grafting success rate is over 93.8% and the system has a working capability of 327 seedlings per hour; the overall adhesion rate for the grafted seedlings kept in an acclimatization chamber is over 84.4%. Demonstrations were held and farmers were highly satisfied with the system performance.

Estimation of Leaf Nitrogen Content using Artificial Neural Network with Cross-Learning Scheme and Significant Wavelengths

Reflectance from crops provides spectral information for non-destructive monitoring of their nutrition status. In order to develop a multi-spectral imaging system for remote sensing of the nitrogen content of crops, the significant wavelengths and calibration models were carefully evaluated in this study. The significant wavelengths in full band (400-2500 nm) and a selected band (450-950 nm), which is suitable for silicon CCD cameras, were investigated. In this article, significant wavelengths for estimating nitrogen content of cabbage seedling leaves were first determined by SMLR (stepwise multi-linear regression) analysis. A proposed ANN (artificial neural network) model with cross-learning scheme (ANN-CL) was further developed to increase the prediction accuracy. To comply with the design of a practical multi-spectral imaging system using silicon CCD cameras and commercially available bandpass filters, an ANN-CL model with four inputs of spectral absorbance at 490, 570, 600, and 680 nm was developed. The calibration results (rc = 0.93, SEC = 0.873%, and SEV = 0.960%) reduced the SEV about 15% when compared with the SMLR method with four wavelengths (SEV = 1.099%). In addition, the results were comparable to that of SMLR with seven wavelengths (rc = 0.94, SEC = 0.806%, and SEV = 0.993%) in the full band. These results indicated that the ANN model with cross-learning using spectral information at 490, 570, 600, and 680 nm could be used to develop a practical remote sensing system to predict nitrogen content of cabbage seedlings.

Development of a Circular Grafting Robotic System for Watermelon Seedlings

The aim of this research was to develop a top plug-in grafting robotic system that is applicable for grafting a scion into a mature rootstock before the scions cotyledons spread. This would require the scion to be very delicate and greatly different from the rootstock in terms of seedling age. The grafting robotic system for watermelon seedlings consists of a rootstock processing unit and a scion processing unit, with dimensions of 120 cm 105 cm 130 cm. Both the rootstock and scion processing unit have a rotating disc, with the disc of the rootstock processing unit being lower than that of the scion processing unit; furthermore, the two discs rotate in opposite directions. A Geneva Wheel intermittent motion mechanism was adopted to drive the discs so that simultaneous and opposite movement was possible. The rootstock processing unit takes charge of removing the leaf bud from the rootstock seedling and drilling a hole for the scion, while the scion processing unit performs the action of cutting the scion seedling to make a sharp angle and finishes the grafting operation by inserting the scion into the hole in the rootstock. The developed grafting robotic system is characterized by the absence of grafting clips. This research employs bottle gourd Chiang-Li #1 as the rootstock and watermelon Fu-Bao #2 as the scion to conduct a series of mechanical grafting experiments. The experimental results showed that the grafting robotic system was able to accurately finish all grafting procedures and operations, with an average success rate of 95% and a working capability of 480 seedlings per hour. Demonstrations were held and farmers were highly satisfied with the systems functions and its performance.

Optimization of suitable ethanol blend ratio for motorcycle engine using response surface method

In view of energy shortage and air pollution, ethanol-gasoline blended fuel used for motorcycle engine was studied in this work. The emissions of carbon monoxide (CO), nitrogen oxides (NOX) and engine performance of a 125 cc four-stroke motorcycle engine with original carburetor using ethanol-gasoline fuels were investigated. The model of three-variable Box Behnken design (BBD) was used for experimental design, the ethanol blend ratios were prepared at 0, 10, 20 vol%; the speeds of motorcycle were selected as 30, 45, 60 km/h; and the throttle positions were set at 30, 60, 90 %. Both engine performance and air pollutant emissions were then analyzed by response surface method (RSM) to yield optimum operation parameters for tolerable pollutant emissions and maximum engine performance. The RSM optimization analysis indicated that the most suitable ethanol-gasoline blended ratio was found at the range of 3.92–4.12 vol% to yield a comparable fuel conversion efficiency, while considerable reductions of exhaust pollutant emissions of CO (-29 %) and NOX (-12 %) when compared to pure gasoline fuel. This study demonstrated low ethanol-gasoline blended fuels could be used in motorcycle carburetor engines without any modification to keep engine power while reducing exhaust pollutants.

Study of an Autonomous Fruit Picking Robot System in Greenhouses

Abstract The objective of this research was to develop an autonomous picking robot system for greenhouse-grown tomatoes, which consists of four major components: the end-effector, machine vision, robot carrier, and control system. The graphical programming language LabVIEW ver. 7.1 was employed to develop the image processing and control system. The experimental results showed the success rates of the integrated picking were 94.83 %, 91.83 %, and 89.63 %, respectively. The average picking time needs about 35.96 s/sample, with a throughput of 100.1 samples/h. Consequently, an autonomous picking robot system was successfully developed and it needs to be further tested for real tomato picking operations in greenhouses in the future.

Development of multi-functional remote sensing system for greenhouse production

This study is aimed to develop a multi-functional remote sensing system based on spectral imaging and environmental sensing for seedling production in the greenhouses. The spectral images were grabbed with exposure time and signal gain controls through IEEE-1394 interface; and a color camera and a B/W camera with optical filter at 780 nm were used. A control program was developed to grab the good quality images using the automatic exposure algorithm with a developed software using Matlab and LabVIEW. To obtain necessary spectral information regarding tray locations and seedling growth status on greenhouse benches, a serial image processing procedures, including spatial calibration, image stitching, gray-level calibration and image segmentation were developed. The data of tray positions and growth status were transferred to the look up table (LUT) and delivered to the water management module through the DataSocket server and wireless network. Besides, the environmental sensing sub-system, including temperature, relative humidity, and lighting measurements, was also developed with the PCI-6023 interface to analyze the spatial distribution of these parameters in the greenhouse. The information of environmental status will provide a better management for seedling growth in greenhouses.

NEURAL NETWORK ANALYSIS OF ENVIRONMENTAL CONDITIONS INFLUENCING CABBAGE SEEDLING QUALITY

Adequate environmental control for seedling growth is essential in developing a seedling cultivation system. This study focuses on the model development of a neural network model to investigate the relationship between the quality of cabbage seedlings and their growth environment. Three different neural models were developed and evaluated. An important approach adopted in this work is that the seedling growth is considered as a result of the cumulative effects of many interacting influences in the growth process. Thus, a historical growth factor, daily dry matter increase weight in the preceding stage, is included in the model. By integration of schemes for various growth stages and the historical growth factor, the model contributes markedly in prediction ability. The error is decreased by 77% (from 33.7% to 7.87%) when the best model developed in this work is used.

Using Optimization Methodology to Identify Suitable Biodiesel Blend Ratio

Biodiesel is a renewable alternative fuel produced from a wide range of vegetable oils and animal fats for use in compression ignition engines. It is rapidly expanding around the world, making it imperative fully to understand the impacts of biodiesel on the diesel engine combustion process and pollutant formation. Biodiesel can blend with diesel in any ratio, and the air pollutant emissions decreased with biodiesel content in fuels. However, biodiesel blend fuel reduced the diesel engine power output. Blend fuels containing 10, 20, 30, and 40% by volume of soybean biodiesel were used. ECOM-AC gas analyzer was used to measure the concentrations of CO, CO2, and NOx in the exhaust gas. The sampling of exhaust gas was taken in extension section of the exhaust pipe without catalytic converter. The engine experimental results showed these exhaust emissions were reduced for biodiesel blended fuels. However, additive for biodiesel was a slight increase in oxides of nitrogen (NOx) emission. From these statistical models, there are influences of biodiesel blend ratio, engine speed and throttle position. Statistical optimization methodology was used to investigate the mutual interaction between the emissions and engine performance. In this study, the suitable biodiesel blend ratio was about 2 ~ 7 %, it was the balance between exhaust emissions and engine performance.

Study of deacetylation in chitinous materials using near infrared spectroscopy

Chitinous materials are important sources for bio-medical applications, and the process monitoring is one of key factors for the quality control of products. In this study, chitin and chitosan in suspension form were analyzed using near infrared (NIR) spectroscopy. Two models including multiple linear regression (MLR), modified partial least square regression (MPLSR) were adopted for studying the degree of deacetylation (DD) of chitinous materials in order to assure a better quality monitoring and control for chitosan production. During the process of the deacetylation, the real-time measurements of suspension were conducted. The MPLSR model with second derivative spectra in the range of 600-1000 and 1400-1500 nm yielded the best results, which were rc=0.991, SEC=0.019, RESC=1.4%, rp=0.990, SEP=0.022, RSEP=3.4%, RPD=9.4. The NIR measurements of DD status of chitinous suspension could be achieved by using the MLR and MPLSR models developed in this study. It provides great application potentials to the real-time and on-line quality monitoring of deacetylation process for the production of chitosan.