ObjectiveTo address the issue of unstable stubble height during the machine harvesting in the first quarter of ratoon rice, which damages axillary buds and affects the yield of the ratoon season. MethodThis paper proposed an automatic control method for regulating the stubble height of ratoon rice. An online monitoring scheme for the header height above ground was established, primarily utilizing millimeter-wave ranging sensors and displacement sensors. The measurement accuracy of the header height above ground was improved by real-time adjustment of the threshold weights for left and right ranging data of the header via a linear displacement sensor. A raw data processing method for the header height above ground, based on multi-segment thresholds and Kalman filtering, was proposed to solve the problem of inaccurate ranging caused by environmental interference such as field leaves and weeds during the measurement process. A fuzzy PID control model was adopted for the automatic regulation of header height. The effectiveness of the control model was verified through AMESIM and Simulink co-simulation, and field harvesting experiments were conducted.ResultField test results showed that the standard deviation of the real-time monitoring data for header height above ground was 5.6 mm, the mean absolute error was 4.8 mm, and the root mean square error was 6.7 mm. When the header height control system was activated, the standard deviation of the monitored stubble height was consistently less than 12 mm, and the coefficient of variation was consistently less than 0.1. ConclusionThe automatic control system for header stubble height of ratoon rice proposed in this paper demonstrates high control accuracy and qualified stubble height, meeting the requirement for stubble height consistency in the first crop of ratoon rice.

Grape pomace is a major byproduct of winemaking and a rich source of bioactive anthocyanins with potential functional value. This study aimed to optimize anthocyanin extraction from Tannat grape pomace and evaluate its antioxidant and anti-aging activities. Ultrasonic-assisted extraction combined with a Box-Behnken design identified optimal conditions of 51.27 degrees C, 53.46% ethanol, 20.10 min ultrasonication, and a 1:24.05 solid-to-liquid ratio, yielding 186.21 +/- 1.03 mg/100 g (R-2 = 0.9798, p < 0.0001). Tannat Grape Pomace Anthocyanins showed strong antioxidant capacity, with 2,2-Diphenyl-1-picrylhydrazyl scavenging of 89.44% +/- 0.87% at 0.2 mg/mL (IC50 = 0.09 mg/mL) and 2,2 '-Azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) scavenging of 95.83% +/- 0.54% at 0.75 mg/mL (IC50 = 0.26 mg/mL). In Caenorhabditis elegans, TGPA extended lifespan, improved motility, and increased heat and oxidative stress resistance without reducing reproductive capacity. Lifespan is a key indicator of aging. This study holds significant implications for advancing our understanding of the mechanisms underlying lifespan regulation, the connection between aging and disease, as well as the development of anti-aging therapies for humans. In conclusion, these findings indicate that Tannat Grape Pomace Anthocyanins possess promising antioxidant and anti-aging potential and support the sustainable, high-value utilization of grape pomace. This approach directly aligns with the core principles of sustainable agriculture by transforming an agricultural byproduct into a valuable resource.

Coordinated operation between a combine harvester and a grain transporter can markedly enhance harvesting efficiency. The crux of such dual-machine collaboration lies in maintaining precise lateral and longitudinal positioning so that grain discharged from the harvester can be accurately deposited into the transporter. To achieve high-precision cooperative unloading control, we establish a leader–follower geometric model of relative positioning based on the harvester–grain transporter kinematic model and then derive the corresponding lateral and longitudinal error dynamics. Building upon an improved power–exponential reaching law, we design a robust, disturbance-resilient sliding-mode controller (RPSMC) for dual-machine lateral and longitudinal position regulation. To address external disturbances such as slip and skidding inherent in cooperative field operations, an improved nonlinear disturbance observer (NDOB) is developed to estimate environmental perturbations; real-time estimation with feedforward compensation further strengthens disturbance rejection throughout the unloading process. A Simulink–RecurDyn co-simulation platform is constructed, and comparative tests under varying soil conditions and disturbance scenarios demonstrate that, relative to baseline control, the proposed method significantly improves response speed and steady-state accuracy. Repeated field experiments further confirm the effectiveness of the proposed method at a cooperative unloading speed of approximately 1 m/s. After convergence, the maximum longitudinal and lateral errors remained within 0.1 m, while the mean absolute errors were below 0.06 m. The maximum steady-state standard deviations were 0.086 m in the longitudinal direction and 0.079 m in the lateral direction, indicating small fluctuation and good repeatability. Compared with the baseline controllers, the proposed method achieved faster convergence, lower peak and average errors, and stronger steady-state disturbance suppression. These results demonstrate that the proposed approach can reliably maintain the unloading alignment between the harvester outlet and the transporter bin center, providing a practical cooperative unloading solution for automated harvesting.

To address the issue of decreased path tracking accuracy of the tracked grain transporter (TGT) due to uncertain external disturbances during field operations, this paper proposes an anti-disturbance sliding mode control strategy that integrates prescribed performance and nonlinear disturbance observer (PPSMC + NDOB). Based on the standard kinematic model of tracked agricultural vehicles, a kinematic error model is constructed that incorporates unknown external disturbances and nonlinear characteristics. To strictly constrain the maximum overshoot and steady-state error boundaries during the path tracking process, a prescribed performance (PP) method is employed to nonlinearly map the lateral tracking errors, dynamically confining the errors within the time-varying boundaries defined by the prescribed performance function. Furthermore, a robust sliding mode control (SMC) based on a power exponential approach is designed, and an improved nonlinear disturbance observer (NDOB) is innovatively introduced. This observer significantly enhances the system’s disturbance rejection capability by real-time estimating external disturbances and performing feedforward compensation. Simulink-RecurDyn co-simulations and field tests show that the proposed control strategy can effectively observe and suppress the influence of external disturbances, and it ensures that the TGT accurately tracks the desired path with prescribed performance indicators (steady-state error band < 0.1 m). Compared to nonsingular terminal sliding-mode control (NTSMC) and Fuzzy pure tracking (FPT) without PP and NDOB in field tests, at the speed of 1.3 m/s, the PPSMC + NDOB reduces the average lateral error by 42% to below 0.05 m, and the standard deviation of the steady-state lateral error is decreased to within 0.1 m, while effectively shortening convergence time and suppressing oscillations in the control signals. This research provides an effective solution for high-precision path tracking control of tracked agricultural machinery in complex farmland environments, holding significant engineering application value.

Agricultural harvesting manipulators are key components of smart agricultural harvesting systems, where collision-free and low-damage motions must be generated under irregular crop distribution, foliage occlusion, illumination variation, deformable obstacles, and uncertain target or obstacle boundaries. In such environments, path planning is influenced not only by geometric feasibility, but also by visibility, sensing reliability, contact safety, and execution stability. This review surveys path planning studies for agricultural harvesting manipulators from the perspective of optimization objectives and methodological development. Unlike crop-specific harvesting robot reviews or broad agricultural robot path-planning surveys, this paper focuses on manipulator-level motion generation across harvesting scenarios. Classical optimization-based methods are first summarized, including single-objective formulations focused on time, path length, energy consumption, smoothness, or safety, as well as multi-objective approaches that address trade-offs among competing criteria. These methods provide the main analytical foundation for agricultural manipulator path planning, but their performance often depends on manually designed objectives, weight settings, and relatively explicit environmental representations, limiting adaptability under changing field conditions. The review then discusses the growing use of learning-based methods, with particular attention to deep reinforcement learning and active learning. These methods are considered mainly as extensions to classical planning, especially for improving adaptability, sample efficiency, and deployment under uncertainty, rather than as direct replacements for model-based approaches. Finally, open challenges are discussed, including contact-aware manipulation, uncertainty-aware planning, task-level robustness, perception-planning-control coupling, and reproducible evaluation for smart agricultural harvesting systems. This review aims to clarify how optimization objectives, environmental uncertainty, and learning-based adaptation can support more robust and deployable path planning for agricultural harvesting manipulators.

The yield assessment process during maize harvesting is a necessary means to ensure farmers’ economic benefits and stable agricultural production. Predicting the mass of maize kernels is an important condition for yield detection. This study proposes a maize kernel mass prediction model based on machine vision and machine learning algorithms to determine whether the kernels are broken. By extracting the geometric features of maize kernels, a phenotypic feature dataset of maize kernels was constructed. Subsequently, popular machine learning algorithms were used to establish regression models for maize kernel mass, achieving quantitative prediction of maize kernel mass. The results indicate that the PLSR (Partial Least Squares Regression) and RF (Random Forest) algorithms are suitable for constructing mass prediction models for broken and unbroken kernels, respectively. The models established by the two algorithms achieved R-values of 0.941 and 0.925, respectively. Field trial results show that there is a strong linear relationship between the predicted maize kernel mass using the constructed model and the actual kernel mass. Therefore, this method can serve as an accurate, objective, and efficient detection method for maize yield.

Complex harvesting environments and varying crop conditions often lead to threshing cylinder blockage and increased entrainment loss in maize grain harvesters. To address these issues, an electric-driven automatic control system for maize threshing concave clearance based on real-time entrainment loss monitoring was developed. The system automatically adjusts concave clearance parameters at different harvesting speeds to maintain grain entrainment loss within an optimal range. First, an adjustable concave structure based on a crank-link mechanism was designed, with a threshing clearance adjustment range of 15–47 mm and motor rotation angle of 0–48°. Subsequently, an EDEM simulation model of the mixed material discharge inside the threshing cylinder was established to determine the optimal installation position of the entrainment loss monitoring sensor based on piezoelectric ceramic-sensitive elements. The sensor was positioned at the left tail end of the concave sieve, with a minimum distance of 58 mm between the sensitive plate centerline and threshing concave sieve and an installation angle of 65° relative to the horizontal plane. A maize threshing clearance control method based on fuzzy neural network PID control algorithm was proposed, and Simulink simulation optimization verified its superior performance with fast response speed. After system integration, field trials were conducted at low, medium, and high operating speeds with preset ideal entrainment loss intervals. The results showed that control was unnecessary at low speed, the control system-maintained entrainment loss within set range at medium speed, and maximum threshing clearance was needed at high speed. Finally, comparative trials of threshing performance with and without the control system were conducted at medium harvesting speed. Results showed that the entrainment loss rate decreased by 43.75% with the control system activated, significantly reducing maize threshing entrainment losses. This study overcame the barrier of maize threshing parameter adjustment being heavily reliant on manual experience and provided theoretical support for the intelligent grain harvesting equipment.

To address the issues of leveling difficulties and poor stability of crawler combine harvesters in hilly and mountainous regions, this research analyzed the mechanical causes of overturning instability in crawler combine harvesters and designed an omnidirectional attitude adjustment chassis based on a five-bar mechanism. A 3D model was developed in SolidWorks, and coupled rigid-flexible simulations were performed using RecurDyn. Results showed that the chassis could achieve an overall lift, lateral adjustments and longitudinal adjustments (0-100 mm, -5.18° to 5.55° and -4.06° to 5.15° respectively), with maximum dynamic stress occurring on the left front and left rear rotational arms. A dynamic stress testing system was established to conduct response surface experiments. Field test results revealed that the primary factors affecting the maximum stress of the left front rotational arm were the grain tank loading mass, lateral adjustment angle, and longitudinal adjustment angle. For the left rear rotational arm, the order was the longitudinal adjustment angle, lateral adjustment angle, and grain tank loading mass. Validation tests showed that at a lateral adjustment angle of 3.61°, a longitudinal adjustment angle of 3.20°, and a grain tank load of 350 kg, the average maximum stresses were 483.19 MPa for the left front rotational arm and 188.95 MPa for the left rear rotational arm, with corresponding structural safety factors of 1.61 and 4.31, meeting strength requirements. This work provides methods for optimizing the design and reliability testing of agricultural machinery chassis with attitude adjustment functions in hilly terrains.

To address the issues of signal loss and insufficient accuracy of traditional GNSS (Global Navigation Satellite System) navigation in agricultural machinery sheds and farm access road environments, this paper proposes a high-precision mapping method for such complex environments and a real-time localization system for agricultural vehicles. First, an autonomous navigation system was developed by integrating multi-sensor data from LiDAR (Light Laser Detection and Ranging), GNSS, and IMU (Inertial Measurement Unit), with functional modules for mapping, localization, planning, and control implemented within the ROS (Robot Operating System) framework. Second, an improved LeGO-LOAM algorithm is introduced for constructing maps of machinery sheds and farm access roads. The mapping accuracy is enhanced through reflectivity filtering, ground constraint optimization, and ScanContext-based loop closure detection. Finally, a localization method combining NDT (Normal Distribution Transform), IMU, and a UKF (Unscented Kalman Filter) is proposed for tracked grain transport vehicles. The UKF and IMU measurements are used to predict the vehicle state, while the NDT algorithm provides pose estimates for state update, yielding a fused and more accurate pose estimate. Experimental results demonstrate that the proposed mapping method reduces APE (absolute pose error) by 79.99% and 49.04% in the machinery sheds and farm access roads environments, respectively, indicating a significant improvement over conventional methods. The real-time localization module achieves an average processing time of 26.49 ms with an average error of 3.97 cm, enhancing localization accuracy without compromising output frequency. This study provides technical support for fully autonomous operation of agricultural machinery.

The current mainstream text steganalysis methods are based on supervised training, requiring a sufficient number of training samples, and necessitating that the training and testing distributions meet the prerequisite of being independently and identically distributed (i.i.d.). However, in practical applications, the distribution of stego text may vary with factors such as text corpora, steganographic algorithms, and embedding rates. This issue, known as domain mismatch, will significantly impact the models’ detection performance. To address this issue, we propose a multi-task few-shot text steganalysis model based on Context-sensitive Prototypes, namely CP-Stega. CP-Stega first extracts generic and task-specific text features from the input sentences and constructs a prototype for each task by averaging the sentence representations of the support set. Next, CP-Stega measures the similarity between the query sample and each task by calculating the distance between the sentence representations and each prototype for any given query sample. To improve the model's adaptability to diverse tasks, we design a loss function consisting of multi-margin loss and KL divergence loss, which expands inter-task prototype distance and shortens intra-task distance, enabling the model to update its meta-parameters accordingly. Experiments reveal that our model can effectively adapt to diverse tasks under different meta-learning scenarios based on different embedding algorithms and embedding payloads, significantly improving text steganalysis performance compared to current foremost steganalysis models and other meta-learning algorithms. By calculating the distance distribution between the sentence representations and each prototype for any given query sample, CP-Stega effectively measures the similarity between the query sample and each task. To enhance the model's adaptability to diverse tasks, we design a loss function comprising multi-margin loss and KL divergence loss. The experiments demonstrate that CP-Stega can effectively perform various steganalysis detection tasks on three self-built meta datasets with different embedding algorithms and embedding capacities.

Due to the complex unstructured environmental factors in ridge-planting strawberry cultivation, automated harvesting remains a significant challenge. This paper presents an oriented-ridge double-arm cooperative harvesting robot designed for this cultivation. The robot is equipped with a novel non-destructive harvesting end-effector and two self-developed specialized manipulators, integrated with the strawberry picking point visual perception system based on the lightweight Mask R-CNN and a CAN bus-based machine control system. The greenhouse harvesting experiments show that the robot achieved an average harvesting success rate of 49.30% in natural environments after flower and fruit thinning, while only a 30.23% success rate was achieved in untrimmed natural environments. This indicates that the agronomic practice of flower and fruit thinning can significantly simplify the automated harvesting environment and improve harvesting performance. Automated harvesting efficiency test results show that the single-arm average harvesting speed is 7 s per fruit, while double-arm cooperative harvesting can achieve 4 s per fruit. Future expansion by increasing the number of robotic arms could significantly improve harvesting efficiency. However, the study conducted for this paper was poor for those strawberries whose body or stem was severely blocked, which should be further improved upon in follow-up studies.

Traditional track-driven rice combine harvesters, during the grain unloading process, often depend on the operator's frequent adjustment of the grain unloader's position while closely monitoring the accumulation of grain within the truck. Because of the structural characteristics of the harvester and the narrowness of the rural terrain, visibility is often obstructed, thereby increasing the difficulty of operation. It is estimated that the time consumed for unloading grain constitutes almost half of the total harvesting time, significantly reducing the efficiency of the harvester. To address these challenges, this study initially proposes an automated grain unloading system for track-driven rice combine harvesters based on stereo vision and details its operational process. Subsequently, we designed a method for acquiring the expected unloading points based on instance segmentation and another method for acquiring the actual unloading points based on geometric location information. Furthermore, this study introduces a method for determining the depth of the grain unloading truck's frame based on the angle of elevation of the unloading tube and proposes a method for acquiring unloading times at each expected point based on a model mapping grain pile height to unloading time. The experimental results confirm the high stability and reliability of the proposed automated unloading system. The maximum relative error between the computed expected unloading points and the actual expected points is less than 4 %. The maximum relative error between the actual unloading points and the true values is also less than 4 %. The mean absolute error between the obtained depth of the grain unloading truck's frame and the actual value is 0.014 m, with a root mean square error of 0.017 m. At each expected unloading point, the mean absolute error between the actual and expected grain pile heights is 0.012 m, with a root mean square error of 0.014 m. Overall, this study effectively enhances the efficiency of agricultural harvesting, reduces labor requirements, and provides a strong impetus for the automation and intelligence of combined harvesting machinery.

BACKGROUND:Strawberry, being an important economic crop, requires a large amount of human labor for harvesting operations. Efficient and non-destructive harvesting by strawberry harvesting robots requires the precise location of the picking points. Current algorithms for locating picking points encounter significant issues with location errors and minimal effective information in complex situations. RESULTS:To improve the accuracy of the location of picking points, this study proposes a visual location method based on composite models. This method employs object detection and instance segmentation models to detect fruits and segment peduncles sequentially, thereby enabling the identification of picking points and inclination on the peduncle. Different object detection algorithms and instance segmentation models were validated to explore the optimal model combination, and the Convolutional Block Attention Module (CBAM) was integrated into YOLOv8s-seg to construct YOLOv8s-seg-CBAM. Test results show that the composite model built with YOLOv8s and YOLOv8s-seg-CBAM achieved a peduncle detection accuracy of 86.2%, with an inference time of 30.6 ms per image. CONCLUSION:The picking point visual location method based on YOLOv8s and YOLOv8s-seg-CBAM composite models can better balance accuracy and efficiency and can provide more accurate guidance for automated harvesting. © 2024 Society of Chemical Industry.

Rapid and accurate detection of protein content is essential for ensuring the quality of maize. Near-infrared spectroscopy (NIR) technology faces limitations due to surface effects and sample homogeneity issues when measuring the protein content of whole maize grains. Focusing on maize grain powder can significantly improve the quality of data and the accuracy of model predictions. This study aims to explore a rapid detection method for protein content in maize grain powder based on near-infrared spectroscopy. A method for determining protein content in maize grain powder was established using near-infrared (NIR) reflectance spectra in the 940-1660 nm range. Various preprocessing techniques, including Savitzky-Golay (S-G), multiplicative scatter correction (MSC), standard normal variate (SNV), and the first derivative (1D), were employed to preprocess the raw spectral data. Near-infrared spectral data from different varieties of maize grain powder were collected, and quantitative analysis of protein content was conducted using Partial Least Squares Regression (PLSR), Support Vector Machine (SVM), and Extreme Learning Machine (ELM) models. Feature wavelengths were selected to enhance model accuracy further using the Successive Projections Algorithm (SPA) and Uninformative Variable Elimination (UVE). Experimental results indicated that the PLSR model, preprocessed with 1D + MSC, yielded the best performance, achieving a root mean square error of prediction (RMSEP) of 0.3 g/kg, a correlation coefficient (Rp) of 0.93, and a residual predictive deviation (RPD) of 3. The associated methods and theoretical foundation provide a scientific basis for the quality control and processing of maize.

To address the issue of reduced yield in the second season caused by damaged stubbles resulting from being compressed during the harvesting process of the first season’s ratoon rice, a device for rectifying the compressed stubbles was designed. Utilizing the DEM-MBD coupling simulation method, a simulation analysis was conducted to determine the range of key parameters and verify the feasibility of the solution. Using rotational speed, forward speed, and stubble entry angle as experimental factors and stubble rectification rate and second-season yield as evaluation metrics, a three-factor, three-level Box–Behnken response surface field trial was conducted. The theoretically optimal working parameter combination was found to be a forward speed of 1.4 m/s, device rotational speed of 75 rpm, and stubble entry angle of 39°. Under these conditions, three parallel experiments were performed, resulting in a rectification rate of 90.35% in the mechanically harvested and compressed area and a second-season yield of 2202.64 ± 35 kg/hm2. The deviation from the numerical simulation results of parameter optimization was less than 5%. These findings suggest that the designed stubble rectification device for ratoon rice can meet the requirements of stubble rectification during the first-season harvest of ratoon rice. Furthermore, it provides valuable insights for reducing harvest losses in the first season and further improving the level of mechanized harvesting for ratoon rice.

Most existing grain flow sensors are designed for paddle-type elevators, with limited focus on applications in auger elevators. This paper addresses the yield monitoring needs during rice harvesting operations, specifically targeting auger-based outlets through experimental research. An array-type differential grain flow sensor was developed and an indoor test bench was constructed to evaluate its performance. The study compares the effectiveness of time-domain and frequency-domain differential processing, alongside various filtering methods, for pre-processing the sensor’s raw signals. Additionally, a grain flow regression model was built using the Random Forest algorithm. Experimental results demonstrated that the monitoring errors during field tests ranged from -6.42% to 8.23%, indicating that the sensor met the requirements for rice yield monitoring. This sensor provides valuable data for feed rate detection, speed regulation, and adjustments to the threshing and cleaning systems in combine harvesters, offering significant practical implications for the promotion and development of precision agriculture.

Firmware library programming is an important part for MCU STM32 learning. Engineering template is the foundation of firmware programming. The paper presents how the firmware engineering template of MCU STM32 work, the building and important files of the template to help learners with further learning.

A remnant fertilizer monitoring system utilizing three-dimensional (3D) reconstruction was proposed to detect the amount of remaining fertilizer in the applicator's tank. Bench tests were carried out to compare the performance of four algorithms to estimate the remnant fertilizer amount: fertilizer remnant monitoring biharmonic spline algorithm (V4), natural nearest-neighbor algorithm (Natural), linear algorithm (Linear), cubic algorithm (Cubic). The average relative error for remnant fertilizer monitoring is 7.33% for the Linear algorithm, 7.30% for the Natural algorithm, 5.18% for the Cubic algorithm, and 4.30% for the V4 algorithm. Field tests are conducted at three fertilization rates to compare the performances of the V4 and Cubic algorithms. The average relative error for discharged fertilizer monitoring is 8.64% for the Cubic algorithm, which is 1.91% lower than that of the V4 algorithm. The results show that the Cubic algorithm has the best performance for remnant fertilizer monitoring. The average relative error of remnant fertilizer monitoring is 2.42% for the Cubic algorithm, which is 0.43% lower than that of the V4 algorithm. The response time of the remnant fertilizer monitoring system is 0.26 s. The results demonstrate that the proposed remnant fertilizer monitoring system is highly accurate and suitable for real-time applications.

Aimed at addressing the problems of the existing straw choppers on combine harvesters, such as a large cutting resistance and poor cutting effect, combined with bionic engineering technology and biological characteristics, a bionic model was used to extract the characteristics of the cutting blades of locusta migratoria manilensis’s upper jaw. A 3D point cloud reconstruction and machine vision methods were used to fit the polynomial curve of the blade edge using Matlab 2016. A straw-cutting process was simulated using the discrete element method, and the cutting effect of the bionic blade was verified. Cutting experiments with rice straws were conducted using a physical property tester, and the cutting resistance of straw to bionic blades and general blades was compared. On the whole, the average cutting force of the bionic blades was lower than that of the general blades. The average cutting force of the bionic blade was 18.74~38.23% lower than that of a smooth blade and 1.63~25.23% lower than that of a serrated blade. Similarly, the maximum instantaneous cutting force of the bionic blade was reduced by 2.30~2.89% compared with the general blade, which had a significant drag reduction effect. By comparing the time–force curves of different blades’ cutting processes, it was determined that the drag-reducing effect of the bionic blade lies in shortening the straw rupture time. The larger the contact area between the blade and the straw, the more uniform the cutting morphology of the straw after cutting. Field experiment results indicate that the average power consumption of a straw chopper partially installed with bionic blades was 5.48% lower than one with smooth blades, measured using a wireless torque analysis module. In this research study, the structure of the straw chopper of an existing combine harvester was improved based on the bionic principle, which reduced resistance when cutting crop straw, thus reducing the power consumption required by the straw chopper and improving the effectiveness and stability of the blades.

Strawberry picking robot arm is an important part of strawberry picking operation. The narrow and complex working environment of ridged strawberry is a great challenge for strawberry picking. Aiming at the problems of narrow working space and complex working environment for strawberry picking in ridge farming, a P-R-R-P-R-R strawberry picking arm is proposed to realize automatic picking of strawberries in ridge farming. By establishing the kinematics model of the strawberry picking manipulator, the Jacobian condition number and operability index of the strawberry picking manipulator are solved to analyze the dexterity of the manipulator. In order to analyze the performance of each joint of the strawberry picking manipulator in the picking operation, a dynamic model of the strawberry picking manipulator was established. The strawberry picking arm has no singular position within the set strawberry picking point range, and the operability indexes are all greater than 0.38. According to the growing environment of strawberry, the simulation route of strawberry picking and the picking operation cycle were set and tested by experiments. Compared with the S-curve motion mode, the motion of each joint of the strawberry picking robotic arm is less smooth in the uniform acceleration-uniform speed-uniform deceleration motion mode. However, the peak velocity of each joint in the uniform acceleration-uniform velocity-uniform deceleration motion mode is lower than that in the S-curve motion mode, Taking three-joint as an example, the peak velocity of S-curve is 19.8. per second, higher than that of the uniform acceleration-uniform speed-uniform deceleration motion mode, which is 16.6. per second. Since joint 4 needs to overcome the influence of gravity, its peak force exceeds 15.5 N. It did not reach the farthest point of its own stroke during the picking process, resulting in a low moment of inertia at the three joints to the end which caused the peak moment of joint 3 to range from 0.2 N.M to 0.31 N.M. The peak torque of joint 5 is all under 0.1 N.M. The total power consumption of the strawberry picking robotic arm is 0.1622 J in a single picking operation using the uniform acceleration-uniform speed-uniform deceleration motion mode. In conclusion, the strawberry picking manipulator designed in this study has good dexterity and can adapt to a ridge environment.