Electric tractors driven by dual motors have become a critical research topic in the field of pure electric tractor research. As a key component of the transmission system, the optimization of the internal parameters of the power-coupled transmission gearbox has a crucial impact on the power transmission of the whole machine. This work combines the characteristics of low speed and high torque during tractor operation, and adopts the transmission form of double motor input and double planetary group coupling output to design the transmission structure of the gearbox. Then, this paper proposes a dynamic optimization method of the transmission system based on the Improved Deep Deterministic Policy Gradient (IDDPG) algorithm, which realizes the optimization of the gear ratio of the transmission system by constructing a virtual prototype and hardware-in-the-loop simulation environment. In transport mode, the optimized gear ratios shorten the acceleration time of the tractor from 0-20 km/h by 13.6% and increase the motor efficiency by 10%; in rotary mode, the acceleration performance is improved by 28.5% and the motor efficiency is increased by 5%. The study shows that the proposed method is significantly better than the traditional static design and provides a new technical path for the intelligent optimization of the electric tractor drive train, while promoting the efficient and sustainable development of agricultural machinery.

In order to improve the intelligent degree of the micro tiller and meet the increasing demand of "replacing people with machines", a crawler type automatic navigation electric micro tiller suitable for gardening environment was designed. First, the NVIDIA Jetson Orin nano equipped with the robot operating system (ROS) is used as the main control to establish the navigation control system and communicate with the bottom layer of the electric micro tiller. Secondly, based on Kalman filter and extended Kalman filter, according to the influence of gross error on extended Kalman filter state estimation, an improved robust adaptive extended Kalman filter is proposed to effectively correct single error and multiple special errors, so as to limit and reduce the negative impact of gross error on Beidou guidance PVT (position, velocity, time) solution to the greatest extent, and improve the positioning accuracy and signal stability. The shuttle path planning method was used to carry out the navigation test and record the positioning data, and the positioning error standard deviations of x, y, z axes corresponding to different positioning data processing methods were compared. The results show that the improved robust adaptive extended Kalman filter proposed in this study has certain advantages in improving the positioning accuracy and signal stability. The box Behnken test was designed with vehicle speed, tillage depth and rotary blade speed as test factors and soil fragmentation rate as evaluation index; The test results were analyzed by variance, and the regression model of evaluation index and test factors was established. In order to improve the broken soil rate as the optimization goal, the operation speed, tillage depth and rotary blade speed were optimized. The optimal parameter combination was obtained as operation speed 1.4 km/h tillage depth 6 cm and rotary blade speed 234 r/min. The field tillage experiment showed that the broken soil rate could reach 95.8%, which showed that the crawler type automatic navigation electric micro tiller had better operation quality and could meet the requirements of rotary tillage in horticultural environment.

The behavior of oil sunflower seeds penetrating screen holes is an important factor that affects the screening performance of oil sunflower seeds. In this study, a double-deck reverse-motion vibrating screening device for oil sunflower seed screening was designed. The force condition and motion law of the oil sunflower seeds on the screen surface were analyzed. This study compared the effect of particle filling amount of discrete element model of oil sunflower seeds on the simulation effects. The screening process was numerically simulated using the coupled Discrete Element Method and Multibody Dynamics (DEM-MBD) technique with the screening percentage of oil sunflower seeds as the index. The influence of the operating parameters of the vibrating screen on the screening effect was analyzed using a multiparameter collaborative optimization scheme. The results of this study can provide a reference for the numerical simulation of crop screening behavior and the development of screening devices.

With the rise in global meat consumption and chicken becoming a principal source of white meat, methods for efficiently and accurately determining the freshness of chicken are of increasing importance, since traditional detection methods fail to satisfy modern production needs. A non-destructive method based on machine vision and machine learning technology was proposed for detecting chicken breast freshness. A self-designed machine vision system was first used to collect images of chicken breast samples stored at 4°C for 1-7 d. The Region of Interest (ROI) for each image was then extracted and a total of 1254 ROI images were obtained. Six color features were extracted from two different color spaces RGB (red, green, blue) and HSI (hue, saturation, intensity). Six main Gray Level Co-occurrence Matrix (GLCM) texture feature parameters were also calculated from four directions. Principal Component Analysis (PCA) was used to reduce the dimension of these 30 extracted feature parameters for multiple features image fusion. Four principal components were taken as input and chicken breast freshness level as output. A 10-fold cross-validation was used to partition the dataset. Four machine learning methods, Particle Swarm Optimization–Support Vector Machine (PSO-SVM), Random Forest (RF), Gradient Boosting Decision Tree (GBDT), and Naive Bayes Classifier (NBC), were used to establish a chicken breast freshness level prediction model. Among these, SVM had the best prediction effect with prediction accuracy reaching 0.9867. The results proved the feasibility of using a detection method based on multiple features image fusion and machine learning, providing a theoretical reference for the non-destructive detection of chicken breast freshness.

The high-gap plant protection machine is taken in this paper as the research object to ensure the good driving power and safety of the high-gap plant protection machine, and the control strategy of inter-shaft torque distribution is established under different working conditions to improve vehicle power and lateral stability. The anticipated demand torque is initially determined based on the structural characteristics and operational principles of the plant protection machine. Subsequently, a hierarchical control framework is devised by incorporating a formulated switching control strategy. Finally, a simulation model for torque distribution control strategy between shafts is developed on the Matlab/Simulink platform, followed by simulation and experimental verification. The results are presented as follows: the inter-shaft torque distribution strategy established in this paper increases the average longitudinal acceleration by 0.13 m/s² and 0.14 m/s² under the control of low and high to low adhesion road surfaces, respectively. Under the control of the single-line shifting condition, the yaw velocity can successfully follow the expected value with a maximum value of 0.61 rad/s. The side deflection angle of the center of mass does not exceed 2.8°, which can follow the ideal trajectory and improve power and safety.

The displacement error of intelligent patrol robot will continuously increase when it conducts patrol inspection along a fixed trajectory, thus deviating from the established inspection route. This paper presents a displacement compensation method based on CS-BP neural network. The Cuckoo search algorithm is used to optimize the weight and threshold of BP neural network in order to obtain the most excellent neural network structure. The experimental results show that the compensated displacement curve is closer to the actual displacement curve, and the displacement error is much smaller than the uncompensated theoretical displacement curve. At 58.8 s in the experiment, the deviation between the output displacement and the actual displacement after compensation by CS-BP neural network can be maintained at about 3 cm. The displacement compensation method is feasible and can effectively alleviate the problem that the motion path of the intelligent inspection robot deviates from the inspection route.

With the continuous development of Internet of Things technology, various work scenarios have higher requirements for real-time communication between devices. This paper designs a temperature and humidity monitoring system for garlic cold storage, using ZigBee wireless communication technology and the function of serial port transfering to wireless of ESP8266 to realize the real-time collection of temperature and humidity data of garlic during storage. Through the interconnection of Aliyun database RDS with PC and mobile applications, the record of temperature and humidity information in the whole cycle of garlic cold storage is completed, and the function of real-time notification of cold storage alarm information without time and location restrictions is realized. The field experiment is designed and implemented based on the system, and the experimental results show that the operating performance of the system is stable.