Search

Guava fruit is readily concealed by branches, making it difficult for picking robots to rapidly grip. For the robots to plan collision-free paths, it is crucial to segment branches and fruits. This study investigates a fast and accurate obstacle segmentation network for guava-harvesting robots. At first, to extract feature maps of different levels quickly, Mobilenetv2 is used as a backbone. Afterwards, a feature enhancement module is proposed to fuse multi-level features and recalibrate their channels. On the basis of this, a decoder module is developed, which strengthens the connection between each position in the feature maps using a self-attention network, and outputs a dense segmentation map. Experimental results show that in terms of the mean intersection over union, mean pixel accuracy, and frequency weighted intersection over union, the developed network is 1.83%, 1.60% and 0.43% higher than Mobilenetv2-deeplabv3+, and 3.77%, 2.43% and 1.70% higher than Mobilenetv2-PSPnet; our network achieved an inference speed of 45 frames per second and 35.7 billion floating-point operations per second. To sum up, this network can realize fast and accurate semantic segmentation of obstacles, and provide strong technical and theoretical support for picking robots to avoid obstacles.

Innovative dielectric materials with high-temperature resistance and outstanding dielectric properties have attracted tremendous attention in advanced electronical fields. Polyimide(PI) is considered a promising candidate for the modern electronic industry due to its excellent dielectric properties and comprehensive properties. However, the limited-adjustable range of dielectric constant and the difficulty to obtain a high dielectric constant restrict the application of PI as high dielectric materials. Herein, a novel diamine monomer (2,2'-bis((methylsulfonyl)methyl)-[1,1'-biphenyl]-4,4'-diamine (BSBPA)) containing a rigid biphenyl structure and high dipolar sulfonyl pendant groups is designed for high dielectric polyimides. The rigid biphenyl and polar sulfonyl pendant groups can reasonably optimize the molecular structure and orientational polarization of polyimides to improve their dielectric properties and thermal properties. Moreover, the effect of different bridge linkages on the dielectric properties is studied by using the different dianhydrides. Thus, the PI-BSBPA films especially the DSDA-BSBPA film (DSDA: 3,3',4,4'-diphenylsulfonetetracarboxylic dianhydride) achieve great thermal properties (T

Polymer dielectrics with excellent dielectric properties and energy storage performance under elevated temperature are urgently needed in electrical power systems. Polyetherimide (PEI), which is su...

This paper presents an experimental investigation of the shear performance of RC beams strengthened with near surface mounted (NSM) carbon fibre reinforced polymer (CFRP) prestressed concrete prisms (PCPs). The shear behaviour of strengthened beams can be affected by several design variables. In this research, the effect of the following parameters were considered: the prestress level, inclination and spacing of the CFRP-PCPs, and material type of the prism. The control beam had conventional shear steel reinforcement only while the other seven beams were shear strengthened with CFRP-PCPs by varying design parameters mentioned above. All the beams were tested under monotonic loading until they reached the failure load. The experimental results showed that the NSM CFRP-PCPs strengthening technique improves the shear performance of the beams effectively. The strengthened beams that applied the CFRP-PCPs at an inclination of 45 ° were more efficient in improving the shear capacity compared to vertical CFRP-PCPs. The shear capacity and deformation were enhanced with the increase of prestressing levels of CFRP rods and the decrease of CFRP-PCPs spacing. The failure modes of the strengthened beams were influenced mainly by the spacing and the inclination of the CFRP-PCPs. Moreover, the material type of the prism had little influence on the effectiveness of shear strengthening. The analytical model presented was developed to estimate the shear contribution of NSM CFRP-PCPs and the model was found to predict the shear capacity of the tested beams well.

Manual picking of banana bunch is a dangerous, inefficient and labor-intensive aerial work. To solve this problem, in order to realize the intelligent operation of picking, it is a prerequisite to solve the identification and positioning of banana and handle. Combined with the real-time and high efficiency of yolov3 deep learning network in target detection, this paper uses the framework of keras + tensorflow to build Yolov3 model. The average accuracy of banana stalk and banana string target detection is 88.45% and 97.96% respectively.

A row-following system based on end-to-end learning for an agricultural robot in an apple orchard was developed in this study. Instead of dividing the navigation into multiple traditional subtasks, the designed end-to-end learning method maps images from the camera directly to driving commands, which reduces the complexity of the navigation system. A sample collection method for network training was also proposed, by which the robot could automatically drive and collect data without an operator or remote control. No hand labeling of training samples is required. To improve the network generalization, methods such as batch normalization, dropout, data augmentation, and 10-fold cross-validation were adopted. In addition, internal representations of the network were analyzed, and row-following tests were carried out. Test results showed that the visual navigation system based on end-to-end learning could guide the robot by adjusting its posture according to different scenarios and successfully passing through the tree rows.

To explore the effect of different photoperiods on the growth of Anoectochilus roxburghii at different growth period and to provide theoretical basis for determining the suitable photoperiod of A. roxburghii in artificial standardized cultivation. The experiment was carried out in the closed plant factory with 30 days, 60 days and 90 days seedling of A. roxburghii as experimental material. According to the illumination time from short to long, five photoperiods were set up, which were light 10h, light14h, light 16h, light 18h and light 20h, respectively. The growth indexes such as plant height, stem diameter, leaf area, fresh weight, dry weight, flavonoids content and soluble sugar were determined at different growth period. With the extension of the treatment time, the plant height of A. roxburghii in each treatment group increased as a whole. At the later stage of treatment (90 d), the plant height of each treatment increased with the extension of the photoperiod, and the highest value was observed in the light treatment of 20 h. As the treatment time increased, the difference between the different treatments increased. On the 90 days, the stem diameter, leaf area, fresh weight and dry weight all increased first and then decreased with the extension of the photoperiod, and the largest values were observed in the light treatment of 16 h. The flavonoids content and soluble sugar also increased with the treatment time increased for all the light cycle treatment, and the flavonoids content for 14 h and was the higher at 19.13mg/g and 18.05mg/g. Under the 16 h light treatment, the plant height of the 90 day is relatively high, the stem is the thickest, the leaf area is the largest, and the flavonoids content and soluble sugar content are the relatively high. It is the most suitable light period in the vegetative growth period.

Vegetable oils are increasingly replacing animal fats in diets, but malondialdehyde (MDA), a peroxidation product of these oils, has been regarded as toxic; this necessitated investigation of MDA formation during consumption. This study investigated MDA formation in four vegetable oils during frying French fries (FF) and fried chicken breast meat (FCBM) at 180 °C for 7 h. Results showed that MDA contents were lower in oils used for frying foods than in control oils, mainly because MDA was incorporated into the foods. MDA content was lower in FF, but higher in FCBM, due to the different food components. Model oil and food system analyses yielded similar results. MDA bound the hydrophobic helical structure in starch-based FF, but was exhibited greater reactivity with nucleophilic groups in protein-based FCBM, resulting in stronger interaction with FCBM than with FF. Our results indicated the existence of distinct mechanisms underlying MDA migration in different food matrixes.

In the traditional agricultural wireless sensor networks (WSNs), there is a large amount of redundant data and high latency on critical events (CEs) for data collection systems, which increases the time and energy consumption. In order to overcome these problems, an effective edge computing (EC) enabled data collection approach for CE in smart agriculture is proposed. First, the key features data types (KFDTs) are extracted from the historical dataset to keep the main information on CEs. Next, the KFDTs are selected as the collection data type of the software-defined wireless sensor network (SDWSN). Then, the event types are decided by searching the minimum average variance between the sensing data of active nodes and the average value of the key feature data obtained by EC. Furthermore, the sensing nodes are driven to sense the event-related data with a consideration of latency constraints by the SDWSN servers. A real-world testbed was set up in a smart greenhouse for experimental verification of the proposed approach. The results showed that the proposed approach could reduce the number of needed sensors, sensing time, collection data volume, communication time, and provide the low latency agricultural data collection system. Thus, the proposed approach can improve the efficiency of CE sensing in smart agriculture.

A total of 15 representative organophosphorus pesticides (OPPs), a class of water pollutants causing serious global concerns, have been sensitively quantified by internal extractive electrospray ionization mass spectrometry (iEESI-MS) after enrichment with Fe3O4-ZrO2 innovatively synthesized in our lab by a one step coprecipitation method. For the premium enrichment, the amounts of Fe3O4-ZrO2, pH value, adsorption time, type and volume of desorption solvent, and shaking time of desorption were systematically optimized. Under optimized conditions, the proposed method provided low limits of detection (LODs) of 0.14-16.39 ng L-1 with relative standard deviations (RSDs) of less than 8.7%. A wide linear response range of about 4 orders of magnitude was achieved with linear coefficients (R2) of 0.9921-0.9999 for all the analytes tested. The present method also provided good recoveries (85.4-105.9%) with acceptable precision (RSDs < 7.2%) in spiked environmental water samples. Furthermore, multiple analytes including dimethoate, omethoate, etc. were simultaneously detected in a single sample run, which was accomplished within 1 min, resulting in significantly improved analytical throughput for quantitative analysis of bulk amounts of samples. The experimental results demonstrated that Fe3O4-ZrO2-iEESI-MS provided advantages, including high sensitivity, high speed, and reasonable selectivity for the detection of OPPs, showing potential applications in environmental water sample analysis and environmental science.

At present, precision agriculture and smart agriculture are the hot topics, which are based on the efficient data collection by using wireless sensor networks (WSNs). However, agricultural WSNs are still facing many challenges such as multitasks, data quality, and latency. In this paper, we propose an efficient solution for multiple data collection tasks exploiting edge computing-enabled wireless sensor networks in smart agriculture. First, a novel data collection framework is presented by merging WSN and edge computing. Second, the data collection process is modeled, including a plurality of sensors and tasks. Next, according to each specific task and correlation between task and sensors, on the edge computing server, a double selecting strategy is established to determine the best node and sensor network that fulfills quality of data and data collection time constraints of tasks. Furthermore, a data collection algorithm is designed, based on set values for quality of data. Finally, a simulation environment is constructed where the proposed strategy is applied, and results are analyzed and compared to the traditional methods. According to the comparison results, the proposal outperforms the traditional methods in metrics.

This paper studied the effects of different light qualities on the growth and photosynthesis of Anoectochilusroxburghii. Under uniform temperature, humidity, CO 2 concentration, and photosynthetic photon flux density conditions, the morphological indexes, chlorophyll, photosynthetic rate, stomatal conductance, and transpiration rate of nine kinds of red and blue LED combination light sources were analyzed. The results showed that the 1:1 blue to red light treatment was more conducive to the accumulation of fresh weight of A. roxburghii than other light; these conditions also caused an increase in the overall plant height and stem diameter of A. roxburghii. White light treatment had a significant effect on the accumulation of aboveground organs of A. roxburghii. The net light and rate were the highest under the white light treatment, and were negatively correlated with intercellular CO 2 concentration. Blue light treatment promoted the increase of stomatal conductance and the transpiration rate of A. roxburghii, but the net photosynthetic rate was the lowest compared with other light treatments. The greater the proportion of red light, the lower the chlorophyll content, and the difference was significant. The blue chlorophyll content has a significantly higher chlorophyll content.

Polymer materials have been a popular choice for dielectric capacitors for decades, but their application suffers a great limitation due to their poor heat resistance. Herein, a crosslinkable poly (aryl ether ketone) (DPAEK) is synthesized as the matrix, while boron nitride (BN) nanosheets are functionalized with reactive groups, thus obtaining a kind of crosslinked poly (aryl ether ketone)/boron nitride nanocomposites as well as forming a stable chemical bonding structure at interfaces via Diels-Alder reaction. Owing to the strengthened interfaces, the nanocomposites exhibit high breakdown strength (536 MV m−1, at room temperature) and may maintain excellent dielectric stability at high temperature. In addition, the nanocomposites yield a significantly improved discharged energy density (3.83 J cm−3) and high charge-discharge efficiency (90.6%) at 500 MV m−1 and 150 °C. This work demonstrates that constructing crosslinked nanocomposites may be a viable way to fabricate high temperature dielectrics.

A row-following system based on machine vision for a picking robot was designed in our previous study. However, the visual perception could not provide reliable information during headland turning according to the test results. A complete navigation system for a picking robot working in an orchard needs to support accurate row following and headland turning. To fill this gap, a headland turning method for an autonomous picking robot was developed in this paper. Three steps were executed during headland turning. First, row end was detected based on machine vision. Second, the deviation was further reduced before turning using the designed fast posture adjustment algorithm based on satellite information. Third, a curve path tracking controller was developed for turning control. During the MATLAB simulation and experimental test, different controllers were developed and compared with the designed method. The results show that the designed turning method enabled the robot to converge to the path more quickly and remain on the path with lower radial errors, which eventually led to reductions in time, space, and deviation during headland turning.

For achieving accurate detection of greenhouse planting environment and obtaining agricultural information such as illumination and air conditions, a multi-node greenhouse environment information collection system was designed combining environment information integrated sensor fusion technology and Internet of Thing technology. The experimental results show that the system is stable in operation, and the collected environmental information data has high precision, which can better detect greenhouse plantations and provide reliable environmental information data.

For solving the compatibility problem of the control system to different track vehicle platforms, a track vehicle cross-platform control system is designed, which reduces the development difficulty of track vehicle control system based on different control platforms, avoids repeated development, and also provides an application interface for the subsequent development of other functions. In the design, the hardware and software parts of the control system are constructed with the general multi-processor hardware architecture and μC/OS operating system kernel. For adapting the multi-processor hardware architecture, a multi-processor collaboration management task is designed. At the same time, the hardware interface of the bottom hardware management layer and the module package of the top application layer are added, so that the hardware development and software development are isolated from each other. The control module developed by this method can be portable to the track vehicle of other platforms, which is convenient for the development and application of different control system.

As the development of artificial intelligence continues, mobile vehicle has increasingly wider applications in intelligent patrol, intelligent crop, etc. For meeting the mobile vehicle's demands, the positioning and navigation system of mobile vehicle is established based on SLAM (Simultaneous Localization and Mapping).The system adopts Raspberry Pi as the main controller and STM32 as the sub-controller. The function of SLAM is realized by data combination and coordinate correlation conversion, after which the target point is located on the map and the functions of positioning and investigation is achieved. The experiment shows that the system can effectively establish the map, obtain the moving vehicle's coordinate information and realize its function of automatic investigation.

Quantification of ultra-trace analytes in complex biological samples using micro-solid-phase extraction followed by direct detection with internal extractive electrospray ionization mass spectrometry (μSPE–iEESI–MS) was demonstrated. 1-Hydroxypyrene (1-OHP) and papaverine at attomole levels in human raw urine samples were analyzed under negative and positive ion detection mode, respectively. The μSPE was simply prepared by packing a disposable syringe filter with octadecyl carbon chain (C18)-bonded micro silica particles, which were then treated as the “bulk sample” after the analytes were efficiently enriched by the C18 particles. Under the optimized experimental conditions, the analytes were readily eluted by isopropanol/water (80/20, V/V) at a high voltage of ± 4.0 kV, producing analyte ions under ambient conditions. The limit of detection (LOD) was 0.02 pg/L (9.2 amol) for 1-hydroxypyrene and 0.02 pg/L (5.9 amol) for papaverine. The acceptable linearity (R2 > 0.99), signal stability (RSD ≤ 10.7%), spike recoveries (91–95%), and comparable results for real urine samples were also achieved, opening up possibilities for quantitative analysis of trace compounds (at attomole levels) in complex bio-samples.

In an orchard environment with a complex background and changing light conditions, the banana stalk, fruit, branches, and leaves are very similar in color. The fast and accurate detection and segmentation of a banana stalk are crucial to realize the automatic picking using a banana picking robot. In this paper, a banana stalk segmentation method based on a lightweight multi-feature fusion deep neural network (MFN) is proposed. The proposed network is mainly composed of encoding and decoding networks, in which the sandglass bottleneck design is adopted to alleviate the information a loss in high dimension. In the decoding network, a different sized dilated convolution kernel is used for convolution operation to make the extracted banana stalk features denser. The proposed network is verified by experiments. In the experiments, the detection precision, segmentation accuracy, number of parameters, operation efficiency, and average execution time are used as evaluation metrics, and the proposed network is compared with Resnet_Segnet, Mobilenet_Segnet, and a few other networks. The experimental results show that compared to other networks, the number of network parameters of the proposed network is significantly reduced, the running frame rate is improved, and the average execution time is shortened.

Antifungal activities of crude extractum of Nanshancha Seed Cake (NSC), to inactivate postharvest pathogens were investigated. Highest inhibitory rate was found against C. musae, C. gloeosporioides and C. papaya P.Henn, which was much stronger than that by tea saponin. Compared to tea saponin, effects of NSC extractum was relatively weak and similar on C. gloeosporioides Penzig and P. italicum. In an in vivo study, best controlling effects by NSC extractum was found with banana anthracnose disease development, which showed no inhibitory effects by tea saponin. NSC extractum controlled in vitro C. musae growth through directly inhibiting germination rate and germ tube elongation, and causing distortation, rupture and indentation of C. musae mycelium. In banana fruit subject to C. musae inoculation, higher PAL, POD, GLU and CHT activity was observed in banana fruit treated with crude NSC extractum than that of water control fruits. Current study proved the best controlling effects of crude NSC extractum in C. musae in vitro and in vivo development, which through direct inhibition of C. musae growth and increasing defense system of the banana fruit.

At present, the harvesting process of asparagus still depends on manual work, which is faced with the problems of large labor intensity and labor shortage. At the same time, the growing law of asparagus is random and the growth height is different. This paper presents a mechanized harvesting machine of self-walking asparagus. The punching angle and cutting height of pneumatic punching mechanism and the inclination angle of feeding roller can be adjusted in a certain range, which can adapt to the wide working conditions and harvesting demands.

With increasing recording density in computer hard disks, the flying height of the magnetic head becomes 10 nm or less. The numerical method of the control equations needs to be optimized consistently with the decrease of the flying height. The iterative process is usually divergent with traditional methods which cannot ensure the accuracy of the analysis results and the stability of numerical calculations. Firstly, a new scheme is presented to solve the ultra-thin lubricating control equations in this paper. The equilibrium equation is obtained by taking the Couette flow terms and time term as dominating terms in iterative process. Moreover, the weighting flow factor is introduced for the abrupt change of film thickness in order to reduce calculated deviation. Secondly, the rules for variation of flying pose of the magnetic head with time are obtained by the time domain analysis. The results show that the magnetic head can recover to the equilibrium position after being slightly influenced by outside disturbance. And the dynamic lubrication properties of the head suffer heavily from the vibrations in the height direction and fluctuation in the pitch direction. Finally, the analysis of the sine roughness-passing capacity of the head indicates that the head can surpass small sine waves and the equilibrium position can be retrieved after a period of decaying oscillation. The proposed research ensures the accuracy and stability of the numerical calculation of the control equations, and the dynamic characteristics and flying stability of a magnetic head is studied systematically. It can also be the theoretical basis for the analysis of the flight characteristics of the slider.

Insect pests that attempt to feed on the caterpillar-resistant maize genotype Mp708 encounter a potent, multipronged defense system that thwarts their invasion. First, these plants are on “constant alert” due to constitutively elevated levels of the phytohormone jasmonic acid that signals the plant to activate its defenses. The higher jasmonic acid levels trigger the expression of defense genes prior to herbivore attack so the plants are “primed” and respond with a faster and stronger defense. The second defense is the rapid accumulation of a toxic cysteine protease called Mir1-CP in the maize whorl in response to caterpillar feeding. When caterpillars ingest Mir1-CP, it damages the insect's midgut and retards their growth. In this article, we discuss a third possible defense strategy employed by Mp708. We have shown that foliar caterpillar feeding causes Mir1-CP and defense gene transcripts to accumulate in its roots. We propose that caterpillar feeding aboveground sends a signal belowground via the phloem that results in Mir1-CP accumulation in the roots. We also postulate that the roots serve as a reservoir of Mir1-CP that can be mobilized to the whorl in response to caterpillar assault.

Based on an improved genetic algorithm (IGA) with a new genetic coding mode of master-slave chromosome, a virtual enterprise partners selecting modeling method is presented in this paper with a demonstration of a specific application case analysis.

Main characters of virtual enterprise feasibility were analyzed in the paper at first. Several new modern approaches of virtual enterprise implement, such as critical function structure (CFS), application service provider (ASP) and agile supply chain (ASC), were reviewed respectively. Based above-mentioned methods, critical aspects of virtual enterprise implementation were discussed with a specific application case

Agile supply chain (ASC) based virtual enterprise (VE) with integrated resources of its partners has more advantages in the competitive market. Detail structure and interior procedure of a VE organization are presented. ASC based partner selecting model technologies of VE have been studied with a new modeling method of fuzzy evaluation. An instantiation is illustrated to demonstrate the modeling method.