PNNGS, a multi-convolutional parallel neural network for genomic selection.

Genomic selection (GS) can accomplish breeding faster than phenotypic selection. Improving prediction accuracy is the key to promoting GS. To improve the GS prediction accuracy and stability, we introduce parallel convolution to deep learning for GS and call it a parallel neural network for genomic selection (PNNGS). In PNNGS, information passes through convolutions of different kernel sizes in parallel. The convolutions in each branch are connected with residuals. Four different Lp loss functions train PNNGS. Through experiments, the optimal number of parallel paths for rice, sunflower, wheat, and maize is found to be 4, 6, 4, and 3, respectively. Phenotype prediction is performed on 24 cases through ridge-regression best linear unbiased prediction (RRBLUP), random forests (RF), support vector regression (SVR), deep neural network genomic prediction (DNNGP), and PNNGS. Serial DNNGP and parallel PNNGS outperform the other three algorithms. On average, PNNGS prediction accuracy is 0.031 larger than DNNGP prediction accuracy, indicating that parallelism can improve the GS model. Plants are divided into clusters through principal component analysis (PCA) and K-means clustering algorithms. The sample sizes of different clusters vary greatly, indicating that this is unbalanced data. Through stratified sampling, the prediction stability and accuracy of PNNGS are improved. When the training samples are reduced in small clusters, the prediction accuracy of PNNGS decreases significantly. Increasing the sample size of small clusters is critical to improving the prediction accuracy of GS.

DEKR-SPrior: An Efficient Bottom-Up Keypoint Detection Model for Accurate Pod Phenotyping in Soybean.

The pod and seed counts are important yield-related traits in soybean. High-precision soybean breeders face the major challenge of accurately phenotyping the number of pods and seeds in a high-throughput manner. Recent advances in artificial intelligence, especially deep learning (DL) models, have provided new avenues for high-throughput phenotyping of crop traits with increased precision. However, the available DL models are less effective for phenotyping pods that are densely packed and overlap in in situ soybean plants; thus, accurate phenotyping of the number of pods and seeds in soybean plant is an important challenge. To address this challenge, the present study proposed a bottom-up model, DEKR-SPrior (disentangled keypoint regression with structural prior), for in situ soybean pod phenotyping, which considers soybean pods and seeds analogous to human people and joints, respectively. In particular, we designed a novel structural prior (SPrior) module that utilizes cosine similarity to improve feature discrimination, which is important for differentiating closely located seeds from highly similar seeds. To further enhance the accuracy of pod location, we cropped full-sized images into smaller and high-resolution subimages for analysis. The results on our image datasets revealed that DEKR-SPrior outperformed multiple bottom-up models, viz., Lightweight-OpenPose, OpenPose, HigherHRNet, and DEKR, reducing the mean absolute error from 25.81 (in the original DEKR) to 21.11 (in the DEKR-SPrior) in pod phenotyping. This paper demonstrated the great potential of DEKR-SPrior for plant phenotyping, and we hope that DEKR-SPrior will help future plant phenotyping.

Adaptive fractional-order nonsingular terminal sliding mode control and sequential quadratic programming torque distribution for lateral stability of FWID-EVs with actuator constraints.

This paper proposes a novel sliding mode control (SMC) algorithm for direct yaw moment control of four-wheel independent drive electric vehicles (FWID-EVs). The algorithm integrates adaptive law theory, fractional-order theory, and nonsingular terminal sliding mode reaching law theory to reduce chattering, handle uncertainty, and avoid singularities in the SMC system. A sequential quadratic programming (SQP) method is also proposed to optimize the yaw moment distribution under actuator constraints. The performance of the proposed algorithm is evaluated by a hardware-in-the-loop test with two driving maneuvers and compared with two existing SMC-based schemes together with the cases with the change of vehicle parameters and disturbances. The results demonstrate that the proposed algorithm can eliminate chattering and achieve the best lateral stability as compared with the existing schemes.

Residual networks without pooling layers improve the accuracy of genomic predictions.

KEY MESSAGE: Residual neural network genomic selection is the first GS algorithm to reach 35 layers, and its prediction accuracy surpasses previous algorithms. With the decrease in DNA sequencing costs and the development of deep learning, phenotype prediction accuracy by genomic selection (GS) continues to improve. Residual networks, a widely validated deep learning technique, are introduced to deep learning for GS. Since each locus has a different weighted impact on the phenotype, strided convolutions are more suitable for GS problems than pooling layers. Through the above technological innovations, we propose a GS deep learning algorithm, residual neural network for genomic selection (ResGS). ResGS is the first neural network to reach 35 layers in GS. In 15 cases from four public data, the prediction accuracy of ResGS is higher than that of ridge-regression best linear unbiased prediction, support vector regression, random forest, gradient boosting regressor, and deep neural network genomic prediction in most cases. ResGS performs well in dealing with gene-environment interaction. Phenotypes from other environments are imported into ResGS along with genetic data. The prediction results are much better than just providing genetic data as input, which demonstrates the effectiveness of GS multi-modal learning. Standard deviation is recommended as an auxiliary GS evaluation metric, which could improve the distribution of predicted results. Deep learning for GS, such as ResGS, is becoming more accurate in phenotype prediction.

A Conceptual Strategy toward High-Reliability Metal-Thermoplastic Hybrid Structures Based on a Covalent-Bonding Mechanism.

Metal-thermoplastic hybrid structures have proven their effectiveness to achieve lightweight design concepts in both primary and secondary structural components of advanced aircraft. However, the drastic differences in physical and chemical properties between metal and thermoplastic make it challenging to fabricate high-reliability hybrid structures. Here, a simple and universal strategy to obtain strong hybrid structures thermoplastics is reported by regulating the bonding behavior at metal/thermoplastic interfaces. To achieve such, we first researched and uncovered the bonding mechanism at metal/thermoplastic interfaces by experimental methods and density functional theory (DFT) calculations. The results suggest that the interfacial covalency, which is formed due to the interfacial reaction between high-electronegativity elements of thermoplastics and metallic elements at metal surfaces, dominates the interfacial bonding interaction of metal-thermoplastic hybrid structures. The differences in electronegativity and atomic size between bonding atoms influence the covalent-bond strength and finally control the interfacial reliability of hybrid structures. Based on our covalent-bonding mechanism, the carboxyl functional group (COOH) is specifically grafted on polyetheretherketone (PEEK) by plasma polymerization to increase the density and strength of interfacial covalency and thus fabricate high-reliability hybrid structures between PEEK and A6061-T6 aluminum alloy. Current work provides an in-depth understanding of the bonding mechanism at metal-thermoplastics interfaces, which opens a fascinating direction toward high-reliability metal-thermoplastic hybrid structures.

Improved AET Robust Control for Networked T-S Fuzzy Systems With Asynchronous Constraints.

This article proposes a novel improved adaptive event-triggered (AET) control algorithm for networked Takagi-Sugeno (T-S) fuzzy systems with asynchronous constraints. First, taking the limited bandwidth of the network into consideration, an improved AET mechanism is proposed to save the communication resource. Superior to the existing event-triggered mechanism, the improved AET scheme introduces two adjusting parameters, which further contribute to the economization of the communication resource. Second, with consideration of asynchronous premise variables, a reconstructed approach is applied to synchronize the time scales of membership functions of the fuzzy system and the fuzzy controller. Third, to derive a less conservative sufficient condition for the controller design, a new augmented Lyapunov-Krasovskii functional with event-triggered information and triple integral terms is constructed. Meanwhile, by applying a Bessel-Legendre inequality and extended reciprocally convex matrix inequality together, a new control algorithm is derived with less conservatism. Finally, simulations on a cart-damper-spring system are implemented to evaluate and verify the performance and advantages of the proposed algorithm.

Multi-Objective Sliding Mode Control on Vehicle Cornering Stability with Variable Gear Ratio Actuator-Based Active Front Steering Systems.

Active front steering (AFS) is an emerging technology to improve the vehicle cornering stability by introducing an additional small steering angle to the driver's input. This paper proposes an AFS system with a variable gear ratio steering (VGRS) actuator which is controlled by using the sliding mode control (SMC) strategy to improve the cornering stability of vehicles. In the design of an AFS system, different sensors are considered to measure the vehicle state, and the mechanism of the AFS system is also modelled in detail. Moreover, in order to improve the cornering stability of vehicles, two dependent objectives, namely sideslip angle and yaw rate, are considered together in the design of SMC strategy. By evaluating the cornering performance, Sine with Dwell and accident avoidance tests are conducted, and the simulation results indicate that the proposed SMC strategy is capable of improving the cornering stability of vehicles in practice.

Controlled evaporative self-assembly of poly(acrylic acid) in a confined geometry for fabricating patterned polymer brushes.

A simple yet robust approach was exploited to fabricate large-scaled patterned polymer brushes by combining controlled evaporative self-assembly (CESA) in a confined geometry and self-initiated photografting and photopolymerization (SIPGP). Our method was carried out without any sophisticated instruments, free of lithography, overcoming current difficulties in fabricating polymer patterns by using complex instruments.