Search

Puffiness is one of the dominant postharvest disorders in easy-peeling citrus cultivars. In this study, the structural changes between healthy and puffy Satsuma mandarin were investigated and the potential of using optical methods for disorder detection was explored. To gain more insight in this disorder, the external appearance and internal quality attributes were first compared between healthy and puffy Iwasaki Satsuma mandarins at three harvest times. Although no consistent differences were observed in the appearance of fruits, the soluble solids content and Brix minus acid values in puffy mandarin were found to be higher compared to the corresponding healthy fruit. The structural properties of the flavedo and albedo tissue layer in the peel were quantified from X-ray CT scans. Whilst no differences were observed in the size of the oil glands in the flavedo, the pore size in the albedo of puffy mandarin was found to be larger with later harvest. The bulk optical properties of the intact fruit were estimated from laser scatter images with a metamodel calibrated on optical phantoms. The reduced scattering coefficient (μ s ') for the intact fruit was found to be lower in puffy mandarin relative to healthy fruit. The distinction between healthy and puffy mandarin based on μ s ' was further validated on Goku Wase Satsuma mandarin. The results obtained indicate that healthy and puffy mandarin can be separated well based on their μ s ' at all the selected wavelengths. This provides a basis for the non-destructive optical detection of puffing disorder at an early stage. [Display omitted] • External and internal attributes were compared between healthy and puffy mandarin. • Thickness of both flavedo and albedo tissue layers was lower in puffy mandarin. • Pore size in the albedo of puffy mandarin was larger at a later harvest time. • The μ s ' of the intact fruit was lower in the puffy than the healthy mandarin. [ABSTRACT FROM AUTHOR]

Fibrous forage and biomass crops such as grass and maize are used at large scale as feed for ruminants and feedstock for biogas production. These crops are typically harvested by chopping them in small particles which are light and suitable for partial pneumatic conveying, where the particle trajectory is determined by their aerodynamic properties. While the aerodynamic properties of many grains have been studied, no reports were found for perennial ryegrass. Chopped ryegrass samples of leaf blades and stems were collected in the fields. After imaging the fresh harvested particles to quantify their size and surface area, their terminal velocity was measured with the fluidised-bed method. The relationship between the mass, projected area and moisture content of the stems and blades with their terminal velocities were modelled. In line with theory, mass and moisture content was positively correlated with terminal velocity, while projected area was negatively correlated with it. Smaller terminal velocities were observed for the blades (1.75 m s−1) compared to the stems (4.44 m s−1). The drag coefficients of grass particles were calculated, and for blades it was negatively correlated with the particle Reynold's number lying within Newton's regime (103 < Re < 3 × 105), whilst no significant correlation was found for stems. The average drag coefficient of blades (0.75) was larger than stems (0.54). No significant effect (p-value <0.01) of moisture content on the drag coefficient of grass particles was observed. This confirms that the drag coefficient is a shape parameter, related to formation of airstreams around the particle. • Aerodynamic properties of ryegrass were quantified in a wind tunnel. • The terminal velocity of grass blades was 61% smaller than that of stems. • Terminal velocities can be predicted from mass, area and moisture content (R2 = 0.77). • Grass particles' drag coefficient was calculated and compared to similar materials. • The drag coefficient of grass blades was 39% larger than that of stems. [ABSTRACT FROM AUTHOR]

The driving tasks of combine harvester operators can be described as being very exhausting because of the high environmental temperatures, long working days and time constraints operators have to deal with. In order to lighten the job, automation is being introduced. In this study, the focus lies on the automation of the cleaning unit. A non-linear prediction model for the sieve losses has been established by means of fuzzy modelling techniques. Excessive sieve losses can be predicted by making use of a differential pressure measurement under the upper sieve section and the derived sieve load signal. The latter signal is a measure for the loadings by grain and chaff on the upper sieve and can thus be linked with the aeration level of the upper sieve section. Validation results revealed that the best position for mounting a pressure sensor is at the rear section of the upper sieve. Grain loss kernels will be blown out by the fan when the loadings on the upper sieve are very low, while a high loading on the upper sieve will result in sieve overload and thus very high sieve losses. [Copyright &y& Elsevier]

Operating a combine harvester can be described as very exhausting because of the high environmental temperatures, long working days and time pressures. In order to lighten the job load, automation technologies have been introduced. In this study, the focus is on the automation of the cleaning unit. A non-linear prediction model for the material other than grain (MOG) content in the grain bin has been established by means of fuzzy modelling techniques. Different model structures were evaluated and best results were obtained when the input data space consists of two input variables (fan speed and load on the upper sieve section) and is subdivided into three local linear sub-models. It was shown that the cleaning section settings such as lower and upper sieve openings had a minor effect on the content of MOG in the grain bin when compared to other variables such as fan speed and the loadings by chaff, straw and grain on the upper sieve. High MOG values can be expected when both the loadings on the upper sieve are high and the fan speed is low. High fan speeds will result in lower MOG values, independent of the load on the upper sieve, but they can also give high losses. [Copyright &y& Elsevier]

The cleaning process on a combine harvester is a complex process that is influenced by a wide range of parameters such as machine settings, field and crop-related parameters, etc. Because of the high time pressures combine drivers have to deal with, optimal settings for the cleaning section are usually only estimated once for each crop. As a consequence, differences in temporal and site-specific conditions are neglected. No recent literature is available that considers the interaction between the settings of the cleaning section (like e.g. fan speed, lower sieve opening and upper sieve opening) and the material other than grain (MOG) content in the grain bin, which is, however, an important performance parameter of the cleaning shoe. In this study, a combine harvester was equipped with extra sensors that could contain valuable information necessary to predict the performance of the cleaning section. A non-linear genetic polynomial regression technique was used to rank the pool of potential sensors as possible regression variables for a prediction model of the MOG content in the grain bin. This model is important for the automation of the cleaning shoe. Results showed that the MOG content in the grain bin is influenced non-linearly by differences in the amount of biomass on the sieve section and the fan speed, which are also correlated with each other. [Copyright &y& Elsevier]

The cleaning process on a combine harvester is a complex process, which is influenced by a wide range of parameters such as e.g. machine settings, and field- and crop-related parameters. Because of the pressures on time faced by combine operators, optimal settings for the cleaning section are usually estimated only once for each crop, and as a consequence, differences in temporal- and site-specific conditions are neglected. Previous literature showed the importance of making use of the most appropriate settings to reduce grain losses. In order to gain insight into the cleaning process of a combine harvester, machine behaviour was registered for different machine settings. Different measurement devices were installed on a test combine to extract information from the combine harvesting process. In order to rank the candidate input variables as possible regression variables for the prediction of the sieve losses, an input selection methodology based on the principles of natural selection is presented. Results indicate that that the sieve losses are affected in a non-linear manner by differences in the pressure profile of the cleaning section and the upper sieve opening. [Copyright &y& Elsevier]

In the search for a healthy indoor climate, chemical cleaning agents are replaced by bacillus strains to create an indoor ecosystem in which these harmless bacteria clean the surfaces by consuming the organic molecules in dust and dirt. To optimise these ecosystems, it is important to monitor the bacterial growth and activity, but the reference methods are time consuming and involve a serious time delay. Therefore, the potential of fluorescence spectroscopy for monitoring the activity of bacillus strains was investigated in this study. We confirmed that fluorescence signals could be detected for Bacillus vegetative cells at a load above 108 cells mL−1. As the fluorescence emission was mostly in the 330–400 nm range after excitation in the 280–290 nm range, we could attribute this signal to the tryptophan residues. The Bacillus fluorescence signals could be distinguished from those released by pure l -tryptophan, plate count agar medium, and Bacillus spores, which were classified with 100% accuracy by PLSDA. In comparison to other microbes, bacillus strains, Escherichia coli (MG1655) and Saccharomyces cerevisiae were found to produce an indistinguishable peak at 340 nm in response to excitation at 280 nm. However, using a hybrid multivariate classifier involving PCA combined with a One-Class SVM on the area-normalised fluorescence emission spectra in response to excitation at 280 nm allowed footprints of Bacillus cells to be identified with a classification accuracy of 78%. • Bacillus subtilis can be detected at above 108 cell mL-1. • Fluorescence of bacillus cells is mainly generated by tryptophan residues. • A hybrid multivariate classifier can discriminate Bacillus strains from other microbes. [ABSTRACT FROM AUTHOR]

Driven by increasing food demand and the urgent need for intelligent land use, the size and complexity of agricultural machinery has increased significantly. Monitoring the operation of machines during long working days is a very challenging task for the operators. Automatic monitoring systems can lighten the load and also pave the way towards fully autonomous machines. This paper proposes an automatic condition monitoring system for detecting operation anomalies on agricultural machine components with reciprocating motion, and applies it to a use-case on a combine harvester header cutter bar. Cyclostationary analysis techniques are employed to develop filtering algorithms to extract informative features, which are monitored through the implementation of statistical process control (SPC) using control charts. Together with a comparison of several filtering and feature extraction techniques, an analysis is provided on the influence of sensor type and position on anomaly detection performance. Filtering benefit was found to be highly dependent on the considered sensor type and its location, with increases in Matthews correlation coefficient (MCC) ranging from 0 to 50%, resulting in maximal MCC values of 1. Suitable feature calculation resulted in average prediction performance improvements over 10% in MCC values for nearly all considered sensor types and their locations. These results highlight the importance of intelligent sensor selection for condition monitoring purposes on agricultural machinery and the added value of SPC involving cyclostationary analysis techniques for anomaly detection. • Knife impacts are detected with a Matthews correlation coefficient of 0.995. • Sensors mounted closer to the cutter bar performed best. • The proposed synchronous filters clearly improved performance. • The standard deviation per rotation captured the signal response during impact well. [ABSTRACT FROM AUTHOR]

Ethiopia is well known for its use of an ard plough dating from antiquity – maresha – which fractures and disturbs the soil. However, hardly any notable progress of experimental research on this animal drawn tillage tool in the field has been made. The attendant problems in current practise are soil- maresha interaction, viz., uneven oxen strength along with different pace of walking, uncontrolled implement behaviour, and field conditions. Taking stock of the experimental research on animal drawn tillage tools in general, most of the documented works on the dynamics of the interaction between soil and animal drawn tillage tools tend to rely on trial-and-error based on factors mainly based on experience and cultural context. As such, no research tailored to systematically handle the link between maresha plough and soil bin experiments exists. To this aim, this study developed a mobile in-situ soil bin facility in which the system was calibrated, tested, and evaluated under outdoor experimental conditions, wherein online measurements of draught, speed, and depth of tillage were carried out. The insights and observations gained from the experimentation were discussed and reported in terms of smooth run, overload, cyclic forces, zero speed with minimal force, stoppage, speed measurement with no force, force measurement with no speed, and low speed with low force. [ABSTRACT FROM AUTHOR]

Fault detection and diagnosis in process industry have attracted a lot of attention recently. There is an abundance of literature on process fault diagnosis ranging from analytical methods to artificial and statistical methods. From a modelling perspective, the methods can rely on quantitative, semi-quantitative and qualitative models. At the other end of the spectrum, there are historical data-based methods that do not make use of any form of model information but rely only on historical process data. The basic aim of this study is to emphasize the importance of introducing more advanced multivariate fault diagnostic systems on agricultural machinery. Up till now, farmers and contractors still observe the process in order to detect process and sensor failures which can disturb the actions of the controllers and cause severe damage to the machine. In the future, the complete reliance on human operators for the correct functioning of these systems will become too risky, due to the increasing complexity of this type of machinery. A systematic and comparative study of various fault diagnostic methods, from an agricultural machinery perspective, is provided in this study. The different fault diagnostic techniques, investigated in scientific literature, are compared and evaluated on a common set of criteria. Typical requirements of a fault diagnostic system for agricultural machinery are adaptability to process changes, user-friendliness, quick detection and robustness. Based on these findings, a hybrid framework of qualitative model-based fault detection techniques and pattern recognition-based methods, which rely on historical process data, is proposed as the most suitable fault diagnostic technique. As a first step towards more advanced fault detection and isolation systems, the general applicability of intelligent neural network techniques like supervised self-organizing maps (SOMs) and back-propagation neural networks is illustrated for the detection and isolation of sensor failures on a New Holland CX combine harvester. Pattern recognition techniques, such as neural networks, were found to be very suitable for this kind of application because a lot of historical process data is available since the recent generation of combine harvesters is equipped with a wide range of sensors and actuators, which are continuously monitored. Moreover, these pattern recognition techniques allow quick detection, are easy to use and are able to adapt their structure and/or model parameters based on new measurement data. Since there is room for improvement of these standard techniques, suggestions for future research concerning fault diagnosis on agricultural machinery are given as well. [Copyright &y& Elsevier]

The sugar profile of a syrup determines its suitability for use in different food applications. The traditional wet chemical method for sugar profile analysis is time-consuming, destructive, and requires several chemical reagents. Therefore, this study aimed to develop a rapid method for determining the concentration of the main sugars (glucose, maltose, maltotriose, and fructose) in tapioca syrups using NIR spectroscopy in the 12000–4000 cm−1 wavenumber region or 800–2500 nm wavelength range. As large spectral variation due to water and liquid glucose was observed, orthogonal projection (OP) was evaluated to reduce those effects. Results showed that for all sugars studied, orthogonal projection prior to partial least squares regression (PLSR) improved the prediction performance in terms of the root mean square error of prediction (RMSEP) by 5–47% and the coefficient of determination for prediction (R p 2) by 3–7%. The developed models had high R p 2 and low RMSEP values, which provides opportunities for industrial monitoring of the sugar profiles of tapioca syrup products. • NIR spectroscopy allows quantifying many sugars in tapioca syrups. • Interference removal by orthogonal projection enhances prediction of syrup sugars. • High-accuracy prediction models for the individual sugars were built by using PLSR. • Identification of important wavelengths for syrup sugar profiling was established. [ABSTRACT FROM AUTHOR]

Crop height and density estimation ahead of the combine harvester have been investigated over the last 30 years, but they remain a challenge. LiDAR technology is increasingly being investigated for phenotyping and monitoring of cereals. However, so far, little has been published about the influence of laser mounting position and robust online measurement of height and density from a single LiDAR scan. Therefore, the influence of the angle and height of the LiDAR mounting on crop height and density estimation in wheat and barley was investigated in this study. Tests were conducted in different crop heights, densities, moisture levels and varieties. The crop height was estimated with a root mean squared error of 0.082 m at an angle of 90° between the LiDAR scanning plane and the horizontal. Two methods were compared for crop density estimation. The inter-percentile and the transversal variance method both successfully predicted the crop density, but the variance based method performed better with a coefficient of determination (R2) of 0.77 and a root mean squared error of 82 g [dry ears] m-2. When considering only one variety, the performance improved to reach an R2 of 0.8 and a root mean squared error of 44 g [dry ears] m-2. Variation in mounting angle of the sensor had less effect on the prediction accuracy than the mounting height. • LiDAR allows to estimate crop height and density in small grain crop. • Crop height cannot be estimated properly at angles to the horizontal below 70°. • Variance based crop density estimation is robust against small viewing angle variations. • LiDAR data correlate better to ears dry mass than to number of tillers. [ABSTRACT FROM AUTHOR]

As the grain-cleaning process in a combine harvester relies on pneumatic separation of the grain and chaff, the aerodynamic forces created by the airflow have a large impact on the cleaning performance. Therefore, to optimise a multi-duct cleaning device in a rice combine harvester, First, the equivalent resistances for the different fan ducts were quantified with a resistance model of fluidised grain and an airflow resistance model for the cleaning sieve. Perforated plates with different opening ratios were designed to simulate working loads. A computational fluid dynamics (CFD) model was validated by comparing simulation results with hot-wire anemometer measurements. The effects of working loads on airflow velocity and volume distribution at the ducts, and total pressure distribution inside the fan were studied in simulation and experiments. After integration of the designed multi-duct fan into the cleaning system, airflow measurements inside the cleaning shoe were made and cleaning performance tests were carried out on a test bench. The ideal airflow velocity within the cleaning shoe was investigated by correlating airflow velocities at different measuring points and analysing the cleaning performance. An evaluation of this multi-duct cleaning system in the field showed large improvements in terms of the grain sieve losses ratio and grain impurity ratio compared to a combine harvester with the single-duct cleaning system. The corresponding grain sieve loss decreased from 2.46% to 0.08%, while the grain impurity ratio decreased from 4.78% to 0.511%. • Resistance of working load at each duct is estimated under working condition. • Variation of airflow and pressure in each duct were analyzed in detail. • With the increasing cleaning loads, airflow velocity at each duct was declined. • The error between CFD computed airflow velocities and measurement results ≤ 15%. [ABSTRACT FROM AUTHOR]

Rice is one of the most important grain crops in China and most of these fields are harvested by combine harvesters. With increasing rice grain yields and feed rates, the cleaning systems rice combine harvesters have to deal with larger volumes of threshed output. Field experiments have shown that the standard single-duct cleaning systems used in most rice combine harvesters has become a limiting factor, resulting in high grain losses and producing high grain impurity ratios. To investigate the reasons for poor cleaning performance, the terminal velocity of the different categories of threshed outputs were quantified and turbine flow meter measurements were performed to obtain insights into the airflow distribution inside the cleaning shoe. Based on these insights, a multi-duct cleaning device with return pan was proposed for tangential-longitudinal-flow rice combine harvesters. The new cleaning system was benchmarked against a commercial system during parallel field tests and it was concluded that the design adaptations improved the sieve losses by 85% and the grain impurity ration by 73%. • Terminal velocity of different components of threshing outputs were quantified. • Distribution of the threshing outputs in the cleaning shoe were studied. • The proposed multi-duct cleaning device improved cleaning performance greatly. [ABSTRACT FROM AUTHOR]

For sustainable pork production and maximum pig welfare, all health, welfare and productivity problems in the barn should be detected as early as possible. In this paper, an automated monitoring and warning system is proposed. Based on measurements of the feeding pattern, it is able to generate daily alerts for individual fattening pigs. Using historical data, the following types of warning systems were developed: (1) fixed limits that treat all pigs and all days equally; and (2) time-varying individual limits using the concept of Synergistic Control. These types of limits were constructed either for the number of registrations per pig or the average interval between feeding visits of a pig, leading to four warning systems in total. These warning systems were used to generate alerts during an online validation period. During an entire fattening period, all pigs were individually monitored to establish true alerts, false alerts and missed problems. The best performance was achieved for the Synergistic Control method on the number of registrations, with a sensitivity of 58.0%, specificity of 98.7%, accuracy of 96.7% and precision of 71.1%. Severe problems were detected on average within 1.3 days from the start of the problem. These are promising results that provide a solid basis for the development of a system for individual pigs but further improvements are warranted to make the system more practical. Highlights • Four warnings systems based on individual pig feeding patterns were validated. • Time-varying individual limits seem most promising. • Sensitivity was 58.0%, specificity 98.7%, accuracy 96.7% and precision 71.1%. • Severe problems were detected on average already within 1.3 days. • Further improvements are warranted to make the system more practical. [ABSTRACT FROM AUTHOR]

Hyperspectral reflectance imaging using unpolarised light has high potential for various applications in non-destructive quality evaluation of food and agricultural products. However, one of the most important challenges is the existence of glare in the obtained hyperspectral reflectance images due to specular reflection of the illuminating light on the sample surface. Whilst this can be avoided for flat surfaces by optimising the viewing angle, this is not possible for curved food products. Therefore, cross-polarisation has been implemented in this study to block the specular reflection and exclude it from the acquired hyperspectral reflectance images in the 450–1000 nm range. The added value of this approach has been successfully demonstrated for three glossy agricultural products: aubergine and two apple cultivars, for which the quality of the obtained hyperspectral images, expressed by signal-to-noise ratios, could be improved from 1.1 to 3 times depending on the wavelength regions. [ABSTRACT FROM AUTHOR]

Recognising and accurately locating fruits on a tree is a critical challenge in developing fruit-by-fruit robotic harvesting. Many researchers have investigated the potential of red, green, blue (RGB) colour imaging for this purpose, but have had limited success due to the occlusion of the target fruits by foliage, branches or other fruits as well as due to the non-uniform and unstructured nature of an orchard environment. Recently, novel, cost-effective camera systems have become available which provide both colour (RGB) and three dimensional (3D) shape information. As these have shown potential for 3D perception for robots operating in unstructured environments, the potential of such an RGB-D camera for the detection and localisation of red and bicoloured apples on tree was investigated in this study. Images were acquired with this camera system in fruit orchards under a light shield blocking direct sunlight, and an algorithm to detect and localise red and bicoloured apples based on colour and shape features was developed. When the algorithm was applied to the data acquired in these orchards, 100% of the fully visible apples and 82% of the partially occluded apples were detected correctly. The location estimation error was below 10 mm in all the coordinate axes of the Cartesian space. This high detection and location accuracy and short processing time (below 1 s for simultaneous detection of 20 apples), makes the developed algorithm suitable for implementation in a robotic harvesting system, and for yield estimation and orchard monitoring. [ABSTRACT FROM AUTHOR]

In order to optimise the processing of stem crops, insight into the deformation behaviour of the crop and the interaction between crop and machine is essential. Most research in the area of mechanical and physical properties of crop stems is focused on characterising the agricultural products to the point of failure using mechanical parameters and empirical relations. No studies have been conducted on the processes which lead to failure of stems and on the processes that take place after failure. In this paper it is shown that the bending behaviour of wheat and barley stalks is very similar to that of steel tubes. Two consecutive phases can be distinguished: ovalisation and buckling. During ovalisation the forces on the wall tend to flatten the cross-section. When this process continues the flexural stiffness is reduced until the structure becomes unstable and buckles. The cross-section locally completely flattens. This deformed cross-section offers virtually no resistance to bending. Mechanical models described in literature have successfully been adapted to describe the bending behaviour of crop stalks during both phases ( R 2 > 0.98 for ovalisation and R 2 > 0.97 for buckling). The crop species, growing conditions, stem diameter and wall thickness were found to influence the bending process significantly. The presence of a core-rind structure increases the bending resistance of the stems. [ABSTRACT FROM AUTHOR]

The conventional methods frequently used in Cuba to determine some fertility parameters important for sugarcane production, such as organic matter (OM), available phosphorus (P) and potassium (K2O), are difficult, costly, and time-consuming procedures. This study was undertaken to build and validate Visible/Near Infrared Reflectance (Vis/NIR) calibration models of these parameters at landscape level and within a field, by taking into consideration their correlation coefficients with the OM. The parameters P and K2O, which are not spectrally active in the Vis/NIR range should be better predicted when are highly correlated with OM. Also, the wavelength intervals to simplify this methodology were selected. Samples were air-dried before scanning using a diode array spectrophotometer covering the wavelength range from 399 to 1697 nm. The regression models were built by using the linear multivariate regression method Partial Least Squares (PLS), and the nonlinear multivariate regression methods Support Vector Machines (SVM) and Locally Weighted Regression (LWR). At landscape level the best correlations between soil spectra and OM (0.90 ≤ R2 ≤ 0.93; 0.12 ≤ RMSEP≤0.14) were obtained with LWR, followed by K2O with LWR (0.77 ≤ R2 ≤ 0.79; 3.47 ≤ RMSEP≤3.62), Olsen P (0.69 ≤ R2 ≤ 0.81; 0.27 ≤ RMSEP≤0.35) and Oniani P (0.64 ≤ R2 ≤ 0.65; 3.31 ≤ RMSEP≤3.61) both with SVM. Also, the nonlinear regression models gave the best results within a field. The higher values for OM (R2 = 0.92; RMSEP = 0.14) and Olsen P (0.68 ≤ R2 ≤ 0.83; 0.27 ≤ RMSEP≤0.34) were observed with SVM, while for K2O (0.16 ≤ R2 ≤ 0.63; 5.13 ≤ RMSEP≤5.88), and Oniani P (0.70 ≤ R2 ≤ 0.72; 2.32 ≤ RMSEP≤2.52) were obtained with LWR. The soil fertility parameters studied at landscape level and within a field were best estimated by using nonlinear regression models. [ABSTRACT FROM AUTHOR]

Compression of biological materials facilitates their transport and storage. In agriculture, straw and hay are commonly compressed with extrusion, but this process is highly influenced by changing crop conditions. A mathematical description of the compression profile of fibrous materials would be useful to compare the compression characteristics of different materials and to predict the energy required for compressing it to a certain density. In this study, different biomass modelling techniques have been reviewed to select the most useful crop compression model. It is shown that the selected crop model (Faborode model) is appropriate in describing the crop compression up to a density of 145 kg m−3, dry matter (with R 2 > 0.8). The selected stress-deformation relation (Faborode model) involves two crop parameters which are determined for wheat straw and hay at different moisture contents and particle orientations (random and parallel stacking). These parameters allow for separating the time-dependent and the elastic compression behaviour. The relaxation properties of the samples have been estimated by fitting the stress decay over time. It is shown that a Maxwell model cannot properly describe the relaxation (R 2 > 0.7) while the Peleg model resulted in reasonable fits with high R 2 (>0.8). Although the model of Peleg is accurate only after a relaxation time of 1 s, the parameters can be used in describing the relaxation behaviour. [Copyright &y& Elsevier]

Highlights: [•] NIR methods for non-destructive composition analysis of oranges was investigated. [•] NIR with emission head gave best predictions of mass, colour index and vitamin C. [•] Using an integrating sphere gave best prediction of TSS. [•] NIR with emission head can evaluate external and internal characteristics. [ABSTRACT FROM AUTHOR]

Combine harvesters operate all over the world, harvesting different crops under different environmental conditions. In order to cope with varying soil, climate and crop conditions, it is important to adjust and optimize the internal settings of these machines to the varying conditions. Since many sub-processes (e.g. threshing, cleaning, speed control, header control) have to be continuously optimised or surveyed, automatic control systems are needed to aid the operator in this tedious task. In this study, automation of the cleaning section on a combine harvester was investigated. In preceding studies considerable knowledge about the working principles of the cleaning unit had been gathered using data-based models. Based on experimental data and fuzzy modelling techniques, optimum and non-optimum working conditions were identified. In this paper, a fuzzy control system which combines the knowledge of experienced operators with these data-based models is elaborated. The main goal of the controller for the cleaning section was defined as finding an optimum trade off between acceptable sieve losses and return volume for a desired level of MOG in the grain bin and a certain amount of incoming biomass. The performance and robustness of the control system have been evaluated by field tests during wheat harvest. Promising results were obtained, demonstrating the benefits of an automatic control system for the cleaning section of a combine harvester when environmental conditions rapidly change over time. [Copyright &y& Elsevier]

The amount of material other than grain (MOG) is an important quality parameter in grain products. The amount of chaff and straw (MOG) in the harvested grain depends highly on the combine harvester settings. On-line knowledge of the grain cleanness could be used to adapt the settings of the threshing and cleaning section in order to keep the proportion of MOG below a desired set point. To allow on-line monitoring of this proportion of MOG, a multispectral vision sensor is developed which creates a virtual image with maximal contrast between kernels and MOG to allow classification of each individual pixel. By calculating the proportion of MOG pixels, a measure for the weight percentage MOG in the sample is then obtained. In this paper, the selection of a combination of a few wavebands which contain sufficient information for adequate discrimination between kernels and MOG is described. For this purpose, hyperspectral measurements of pure kernels and chaff and straw from different varieties of wheat have been acquired to construct a training set where the pixels in the acquired images serve as the samples and the intensities at the different wavelengths as the variables. Using this training set, the best combination of five wavebands in the range from 400 to 900nm with respect to the classification of individual pixels in an image as either kernel or MOG has been selected by means of genetic algorithms. This combination of selected wavebands will then be used to build a multispectral vision sensor. [Copyright &y& Elsevier]