Design and experimental study of tillage depth control system for electric rotary tiller based on LADRC | Scientific Reports

Scientific Reports volume 15, Article number: 1740 (2025) Cite this article

669 Accesses

Metrics details

This paper proposes an adaptive real-time tillage depth control system for electric rotary tillers, based on Linear Active Disturbance Rejection Control (LADRC), to improve tillage depth accuracy in tea garden intercropping with soybeans. The tillage depth control system comprises a body posture sensor, a control unit, and a hybrid stepper motor, integrating sensor data to drive the motor and achieve precise depth control. Real-time displacement sensor signals are compared with target values, enabling closed-loop control of the rotary tiller. Field experiments conducted at 0.5 km/h and 0.8 km/h, with preset tillage depths of 80 mm and 100 mm, demonstrated the system’s effectiveness. The average standard deviation of tillage depth for the LADRC system was 3.2 mm, compared to 10.5 mm for fuzzy proportional-integral-derivative (PID) control. The LADRC control reduced the rate of change in tillage depth by 68.9% compared to fuzzy PID. This system effectively mitigates potential deviations during operation, ensuring stable and reliable tillage depth control.

With the development of modern agricultural technology, the level of agricultural mechanization continues to improve. As an important agricultural tillage machine, the electric rotary tiller plays a critical role in enhancing both the efficiency and quality of tillage operations1,2. The tillage depth control of an electric rotary tiller is one of the key factors in ensuring tillage quality, directly affecting soil turnover, planting depth of crops, and the contact between soil and seeds, which in turn impacts crop growth and yield3. Traditional tillage depth control methods largely rely on manual adjustment, which is often imprecise and slow in response, making it difficult to meet the requirements of precision agriculture in modern farming. To improve the accuracy and response speed of tillage depth control, electric rotary tillers generally adopt intelligent control systems, enabling real-time monitoring and dynamic adjustment of tillage depth in complex environments, thus enhancing operation quality and efficiency4,5.

At the present time, the core task of automatic mechanical control technology for electric rotary tillers is the regulation of tillage depth6,7. Xiao Maohua et al. proposed a novel fuzzy PID control strategy for dynamic tillage depth control, with the objective of enhancing the quality and efficiency of rotary tillage operations and improving the operational accuracy of the self-propelled electric mini-tiller8. Kai Hu et al. developed a high-precision tillage depth monitoring model for rotary tillers, employing model identification techniques to derive precise control functions. A fuzzy adaptive PID method was employed to adjust the tillage depth, resulting in a notable enhancement in system response speed and anti-interference capability9. Qi Wang et al. devised a speed and slip ratio switching control system for wheeled electric tractors, employing a sliding mode control algorithm to effectively regulate speed and slip rate in response to varying tillage resistances. This approach has the potential to minimize tillage depth errors, enhance efficiency, and preserve depth uniformity10. Anzhe Wang et al. proposed an improved tillage depth control strategy for complex farmland terrain, using mechanical angle sensors to precisely measure the hydraulic lift arm angle, correlating it with the tillage angle to achieve high-precision control. The Hybrid Extended State Observer Back Stepping Sliding Mode Controller (HESO-Back Stepping SMC), which has been introduced, is capable of effectively estimating unmeasured variables and disturbances. It provides continuous and smooth control signals while reducing chattering. Comprehensive simulations and field tillage tests have demonstrated the precision and reliability of the HESO-Back stepping SMC in complex agricultural environments5. Yeon-Soo Kim et al. designed a real-time tillage depth measurement system specifically for agricultural tractors. By integrating sensor fusion methods, this system provides accurate real-time tillage depth data while simultaneously synchronizing with a traction force measurement system. The collected data were compared with traction force predictions derived from ASABE standard equations, confirming the system’s accuracy and reliability11. Other research teams have achieved precise control of tillage depth through methods such as fuzzy PID control strategies, high-precision monitoring models, and sliding mode control algorithms. Most of the research focuses on field operations, where the control structure of the rotary tiller primarily relies on hydraulic actuators and generally uses traditional control methods12,13,14,15. There is still a gap in the research and technological development of rotary tilling equipment for tea gardens, with key technologies such as tillage depth control accuracy, uniformity of tillage depth, and the accuracy of sensor fusion data yet to be breakthrough. This paper investigates the electric rotary tiller for tea gardens, using electric rotary tilling actuators and applying more advanced control algorithms. These measures can effectively improve the response speed, anti-interference capability, precision, and reliability of the electric rotary tiller during the operation process.

This paper addresses the issues of external disturbances and the complex working environment of electric rotary tillers, proposing an adaptive real-time tillage depth control system based on posture sensors and hybrid stepper motors. The system is designed to remotely set and transmit the initial tillage depth, while the posture sensors provide real-time feedback on the tillage depth status. Signals are filtered and predicted using an extended Kalman filter, and advanced LADRC control technology is applied to drive the hybrid stepper motor for precise rotary tillage operations, thereby ensuring the stability and reliability of the tillage depth control system. Furthermore, a grating sensor is integrated into the system to enable the real-time monitoring of the tillage depth. The adaptive real-time tillage depth control system’s accuracy was tested and validated in real-world tillage conditions. This study aims to introduce an innovative approach for intelligent tillage depth control in electric rotary tillers, contributing to the advancement of agricultural mechanization towards precision agriculture. This study is the first to propose and design a rotary tilling machine for tea garden intercropping with soybeans, utilizing advanced LADRC technology to effectively control plowing accuracy. This provides an innovative method for smart farming in tea gardens and contributes to the advancement of agricultural mechanization toward precision agriculture.

The overall configuration of the electric rotary tiller is illustrated in Fig. 1. The electric rotary tiller is comprised of three primary components: the walking mechanism, the working mechanism, and the electrical control system. The body of the electric rotary tiller is referred to as the walking mechanism. The working mechanism is constituted by the rotary tiller, which is located at the front of the body and is mounted on a lead screw sliding table. The electrical control system includes a hybrid stepper motor, a lead screw sliding table, a rotary tiller lift controller, an electric tiller walking controller, and the rotary tiller motor. The overall dimensions of the electric rototiller are 2303 mm in length, 617 mm in width, and 1206 mm in height, with a traveling motor power of 800 W, a rotary motor for tilling of 1000 W, a 100 W hybrid stepping motor, and a 48 V−50 Ah lithium-polymer power battery.

Overall structural diagram of the electric tiller.

The adaptive real-time tillage depth control system is comprised of three principal units: a rotary tiller posture sensor, a control unit, and a hybrid stepper motor lead screw sliding table. Alterations in the posture of the rotary tiller, when the electric rotary tiller traverses an uneven road, are principally discerned by the rotary tiller posture sensor. An STM32F407 control chip is employed by the control unit, wherein the microprocessor incorporates input data from the sensors and executes extended Kalman filtering and predictive processing. The data is then processed by the LADRC controller, which generates the requisite pulse control signals for transmission to the hybrid stepper motor driver. The up-and-down movement of the rotary tiller blade is driven by the hybrid stepper motor lead screw sliding table. Signals indicating the real-time tillage depth are transmitted back to the control unit by the grating sensor, thereby completing the adaptive real-time tillage depth control process. Precise control of the rotary tillage depth is attained by the adaptive real-time tillage depth control system through the following steps, as illustrated in the workflow in Fig. 2.

Tillage Depth Control System Workflow.

The target tillage depth, which has been preset remotely, is received by the control unit via wireless communication. This information is then transmitted to the rotary tiller control system, whereupon calculations are performed, and pulse signals are subsequently sent to the hybrid stepper motor. As a result, the lead screw sliding table is caused to descend to the set depth. The displacement of the lead screw is monitored in real time by the grating sensor, which is mounted on the guide rail, and is compared with the preset tillage depth signal. This ensures that the desired depth is achieved. If the working environment is uneven, angular changes are detected by the attitude sensor, which is highly sensitive. The z-axis position of the electric rotary tiller is obtained through the integration of signals sent by the rotary tiller’s attitude sensor. The precise z-axis position of the electric rotary tiller is provided by the grating sensor, which is used as the primary observation source for the extended Kalman filter. The attitude sensor is used as an auxiliary observation factor. The high accuracy of the grating sensor and the high-frequency characteristics of the attitude sensor are utilized to make a more accurate tilling position estimation, allowing external disturbances during the movement of the electric rotary tiller, such as vibrations and body deviations, to be addressed. The requisite calculations are performed by the control unit, which converts them into pulse signals and transmits them to the hybrid stepper motor. The hybrid stepper motor drives the lead screw sliding table, thereby controlling the ascent and descent of the rotary tiller and ensuring its continued stability in the desired position. Once the lifting movement has been completed by the hybrid stepper motor, the displacement signals from the grating sensor, which represent the actual measured tillage depth, are collected in real time by the control unit. By comparing the actual tillage depth signals with the preset values, errors are adapted and corrected by the control unit using the LADRC control algorithm, thereby forming a complete closed-loop control process that ensures the precision and stability of the tillage depth.

Traditional tillage depth control methods, including position-based and force-based adjustments, each have their respective advantages and limitations. Force-based adjustment is not susceptible to fluctuations in the tilling terrain; however, it can experience errors during real-time soil resistance detection, leading to inconsistent tillage depth. Position-based adjustment, while unaffected by soil resistance, may produce significant errors when terrain fluctuations are severe. In such cases, additional sensors are needed to prevent inaccuracies. In general, two methods for tillage depth control are used: switch-based adjustment and the control coefficient method. The control coefficient method, which utilizes parameters derived from extensive research and practical experience, achieves superior control performance. Given the relatively stable terrain conditions in tea plantations, where the electric rotary tiller is deployed, position-based adjustment provides a distinct advantage.

Considering the aforementioned factors, this study employs a combined measurement method utilizing posture and grating sensors, significantly reducing vibration and other external disturbances in the electric rotary tiller. An innovative LADRC controller is used to further minimize errors, while a hybrid stepper motor mechanism ensure stable and precise tillage depth control. The rotary tiller posture sensor detects surface unevenness in real time, and the data undergoes preprocessing using the extended Kalman filter algorithm. The desired tillage depth is compared with the actual depth to determine the requisite real-time adjustment, which is computed by the LADRC controller and converted into pulse control signals to drive the stepper motor. The grating sensor continuously monitors the tillage depth in real time, providing feedback to the LADRC controller, which facilitates adaptive real-time adjustment of the tiller depth. The sensor installation locations are shown in Fig. 3, and the control system schematic is presented in Fig. 4.

Schematic diagram of automatic tillage depth control system.

LADRC Regulation Principle Diagram.

A hybrid stepper motor is operated as an open-loop control device, with electrical pulse signals being converted into angular or linear displacement. The motor’s speed and stop position are entirely determined by the frequency and count of the pulse signals. Upon the receipt of each pulse signal, the motor is prompted to rotate by a predetermined angle in the designated direction by the stepper driver. The main parameters of the stepper motor are calculated as follows:

In Eq. (1), \(\:{\theta\:}^{{\prime\:}}\) is used to represent the step angle (°), N is the number of steps of the stepper motor (related to the motor driver), θ is the motor’s working rotation angle (°), \(\:{K}_{0}\) is the number of pulses, l is the stroke of the sliding table (mm), P is the lead screw pitch (mm), n is the motor speed (rpm), v is the sliding table movement speed (mm/s), and f is the pulse signal frequency.

When the electric rotary tiller encounters uneven road, positional shifts may occur during operation. Real-time adjustments are made to the stepper motor by the controller based on the position changes detected by the posture sensor, thereby enabling the system to adapt to alterations in the tillage depth direction.

In Eq. (2), ΔK is used to represent the number of altered pulse signals, and ΔL denotes the actual adjusted tillage depth achieved by the stepper motor, which is determined from the pitch and roll angle deviations detected by the tillage depth posture sensor and calculated accordingly. Once the stepper motor has been adjusted in accordance with ΔK, the actual working depth of the rotary tiller is considered to be \(\:{L}_{w}\), while the actual depth detected by the grating sensor is \(\:{L}_{a}\). The real-time adjustment error in tillage depth is subsequently obtained, as illustrated in Eq. (3).

The use of a feedback controller to correct the rotary tiller depth in real time enables the shortest possible approach to zero for \(\:{L}_{s}\). This method, which adjusts the tillage depth in real time based on surface irregularities, has been demonstrated to be highly effective in farming environments where terrain changes are minimal. The posture sensor on the rotary tiller is capable of measuring the surface unevenness of the tillage environment in real time. Let it be assumed that the initial pitch angle is α, and that the pitch angle detected at the subsequent moment is α’. In this case, the distance between the center of gravity O of the tiller and the center of gravity P of the rotary tiller is denoted as x, as illustrated in Fig. 5.

Geometric plot of tillage depth versus attitude angle.

The change in rotary tiller depth z’ following a modification in pitch angle can be calculated according to the principles of geometry, as illustrated in Eq. (4).

In addition to pitch angle affecting changes in rotary tiller depth, the roll angle also exerts an influence on tillage depth, as illustrated in Fig. 5. The impact of a modification in roll angle on the rotary tiller depth is represented by z'', as illustrated in Eq. (5).

Where y is the width of the electric rotary tiller, and β is the roll angle measured by the posture sensor. By combining Eqs. (4) and (5), the real-time variation tillage depth ΔL of the electric rotary tiller can be obtained, as demonstrated by the following equation:

The two-phase hybrid stepper motor discussed in this paper is composed of a stator and a rotor. The magnetic poles of the stator generate a magnetic field that attracts the rotor’s poles, causing the rotor’s magnetic field to rotate in synchrony with the stator’s magnetic field16,17. Both the stator and the rotor are made of permanent magnets. The stator is comprised of eight evenly distributed magnetic poles, which are arranged in two distinct phases. Each phase contains four poles, resulting in a total of eight poles distributed across the stator18. In formulating the mathematical model of the stepper motor, the leakage flux between the stator poles and the rotor’s permanent magnet was neglected, as were the effects of hysteresis, eddy currents, and harmonic components induced in the stator coils. The mathematical model for each phase was developed using the electromagnetic circuit model. As shown in Fig. 6.

Single-phase equivalent circuit for two-phase hybrid stepper motor.

The pulse input signal is denoted as θi, and after being received by the stepper motor, it causes the motor to rotate, with the actual rotation angle denoted as θo. The voltage balance equation for the two-phase hybrid stepper motor is given by19:

Where \(\:{U}_{a}\) and \(\:{U}_{b}\) represent the phase voltages of the two-phase windings, while \(\:{i}_{a}\) and \(\:{i}_{b}\) denote the currents of the two-phase windings. L denotes the average self-inductance, and θ represents the angular displacement of the stepper motor. R represents the phase resistance of the two-phase windings. In this context, J represents the moment of inertia, B denotes the viscous damping coefficient, \(\:{T}_{E}\) signifies the electromagnetic torque, \(\:{T}_{L}\) stands for the load torque, \(\:{Z}_{r}\) is the number of teeth of the stepper motor, and \(\:{K}_{m}\) is the stepper motor constant. In accordance with Eq. (7) and the dynamic characteristics of the stepper motor, assuming single-phase excitation and a zero load torque, the following equation can be derived upon energizing the motor:

At the initial moment, the stepper motor is in equilibrium, and the rate of change of the stepper angle is approximately zero. The equation of motion for the stepper angle increment can therefore be expressed as follows:

where \(\:\delta\:\theta\:\) represents the angular increment, which can be expressed as \(\:\delta\:\theta\:={\theta\:}_{0}-{\theta\:}_{i}\). Since \(\:\delta\:\theta\:\) is extremely small, its linearized infinitesimal polarity can be derived.

The Laplace transform can be applied simultaneously to both sides of Eq. (10) above, with the initial value set to zero, resulting in:

The transfer function of the two-phase hybrid stepper motor is given by the following expression:

The parameters of the hybrid stepper motor used in this study are presented in Table 1.

By substituting the aforementioned parameters into the transfer function in Eq. (12), the resulting expression is as follows:.

In 1999, Jingqing Han introduced an advanced control strategy, known as Active Disturbance Rejection Control (ADRC)20. The fundamental premise of this approach is to conceptualize the inherent uncertainties of the system, both internal and external, as a unified disturbance. This disturbance is then estimated and compensated for through the use of an Extended State Observer (ESO). The core structure of ADRC comprises three key components: the Tracking Differentiator (TD), the Extended State Observer (ESO), and the Nonlinear State Error Feedback (NLSEF) control law. The TD extracts the continuous signal and its derivative from the system input, while the ESO estimates the system state and the total disturbance. Based on the ESO’s output, the NLSEF formulates the control law, enabling precise control of the system state21. The basic configuration is shown in Fig. 7.

The basic structure of the ADRC.

In this context, the variable “v” denotes the input signal, “d” represents the total external disturbance, and “y” represents the actual output.

Due to the nonlinear nature of NLSEF and the associated difficulties in analyzing it in practical applications, it is often replaced by LADRC (Linear Active Disturbance Rejection Control) in engineering. Gao Zhiqiang et al. further refined and optimized LADRC by using linear gains instead of nonlinear gains, thereby simplifying the implementation and turning of the control algorithm22,23. The simplified LADRC structure is shown in Fig. 8.

The basic structure of the LADRC.

In this context, the variable “r” represents the input signal to the LADRC. As shown in the basic structure diagram of the LADRC, the TD, LESO, and LSEF components work together to effectively regulate the system. The TD component defines the desired dynamic behavior, the LESO component provides real-time estimates of the system’s state, and the LSEF component generates appropriate control signals based on this information, thereby achieving precise control of the system’s dynamics. This method is particularly well-suited for systems with unknown or evolving dynamics, as it does not require an accurate system model.

As illustrated by Eq. (13), the control object for the rotary tiller of the electric tiller can be identified as a second-order model. Accordingly, the LADRC controller for a second-order object can be designed as follows:

In Eq. (14), u represents the system input, y denotes the output, w signifies the external disturbance, \(\:{a}_{1}\) and \(\:{a}_{2}\) are the system parameters, and b is the control gain. The parameters \(\:{a}_{1}\), \(\:{a}_{2}\), and b are unknown, with \(\:{b}_{0}\approx\:b\). Let \(\:{x}_{1}=y\) and \(\:{x}_{2}=\dot{y}\). The following assumptions are made:

Equation (15) is defined as the generalized disturbance of the system, which includes both internal uncertainties and external disturbances. It is expanded into the system’s state variable, \(\:{x}_{3}=f(y,\dot{y},w)\), from which the state equation of Eq. (16) is derived.

In this equation, \(\:{x}_{1}\), \(\:{x}_{2}\), and \(\:{x}_{3}\) represent the system’s state variables, while h is defined as \(\:\dot{f}(y,\dot{y},w)\). In light of the aforementioned considerations, a linear extended state observer (LESO) is thus established.

By selecting appropriate observer gains \(\:{\beta\:}_{1}\), \(\:{\beta\:}_{2}\), and \(\:{\beta\:}_{3}\), the LESO is capable of achieving real-time tracking of each variable in Eq. (14), that is to say, \(\:{z}_{1}\to\:y\), \(\:{z}_{2}\to\:\dot{y}\), \(\:{z}_{3}\to\:f(y,\dot{y},w)\). Let u be defined as follows: \(\:u=\frac{-{z}_{3}+{u}_{0}}{{b}_{0}}\). Furthermore, let us neglect the estimation error of z₃ on \(\:f(y,\dot{y},w)\). In this case, Eq. (14) can be simplified into a double integrator cascade structure.

Design of the PD controller:

In this equation, v represents the reference signal, while \(\:{k}_{\text{p}}\) and \(\:{k}_{\text{d}}\) denote the controller gains. In accordance with the principles set forth in Eqs. (18) and (19), the closed-loop transfer function of the system can be derived.

The appropriate selection of gains (\(\:{k}_{\text{p}}\) and \(\:{k}_{\text{d}}\)) can serve to stabilize the system. Moreover, the characteristic equation of LESO is given by the following expression:

By selecting the optimal characteristic equation, \(\:\lambda\:\left(s\right)=(s+{\omega\:}_{\text{o}}{)}^{3}\), the result is:

In this equation, the term \(\:{\omega\:}_{\text{o}}\) is defined as the observer bandwidth. Similarly, the parameters of Eq. (20) may be selected as follows:

In the Eq. (23), \(\:{\omega\:}_{\text{c}}\) is referred to as the controller bandwidth, while \(\:\xi\:\) represents the damping ratio. When the value of \(\:\xi\:\) is set to 1, the configuration of the LADRC control parameters is simplified to the selection of three key parameters: the observer bandwidth \(\:{\omega\:}_{\text{o}}\), the controller bandwidth \(\:{\omega\:}_{\text{c}}\), and the control gain \(\:{b}_{0}\). This significantly simplifies the controller parameter configuration.

The primary function of the LADRC is to convert feedback data from the grating sensor into pulse control signals that regulate the operation of the stepper motor. The conversion process is crucial for achieving precise control in the real-time adaptive tillage depth system. During operation, the grating sensor continuously monitors the actual tillage depth of the rotary tiller, generating corresponding measurements. These measurements are then transmitted to the control subsystem, where they are compared with the preset desired tillage depth. As a result of this comparison, the system generates two deviation signals, denoted as e1 and e2, which represent the discrepancy between the actual and desired tillage depth values24.

LADRC utilizes linear state error feedback control laws to enable real-time adaptive adjustments to the hybrid stepper motor on the rotary tiller sliding platform, based on the deviation signals, thus allowing precise regulation of the tillage depth variations. The adaptive adjustment mechanism allows the rotary tiller system to more effectively mitigate the effects of uneven surfaces, shocks, and vibrations from the tilling environment, thereby providing more stable and precise control.

A LADRC control model was developed in Simulink using Matlab 2023a software, based on the adaptive LADRC control method for tillage depth. The resulting model is shown in Fig. 9. In the simulation, step response experiments were conducted at preset tillage depths of 80 mm and 100 mm. The step responses are presented in Fig. 10.

Simulation model of the LADRC controller.

Step response plots for different tillage depths.

In general, the overshoot of traditional controllers such as fuzzy PID is below 20%, and only a few controllers can achieve performance with no overshoot. Although there are performance differences between various rotary tiller controllers, the overshoot is generally designed to be between 5% and 20%, with the exact value adjusted based on the system’s dynamic characteristics and control requirements9. As illustrated in Fig. 10, the LADRC controller demonstrates superior speed performance in comparison to the fuzzy PID controller. The adjustment time is reduced by 46.4% in comparison to the fuzzy PID controller, and the LADRC controller demonstrates the absence of overshoot. A comprehensive comparison of the results is presented in Table 2.

As shown in Fig. 11, the random road unevenness is characterized by a maximum deviation of 50.0 mm above the reference tillage depth and a minimum deviation of −18.0 mm below it. Simulation experiments were conducted at preset tillage depths of 80 mm and 100 mm, with the control results compared to those of the fuzzy PID controller (see Fig. 12). As illustrated in Fig. 12, the tillage depth regulated by the LADRC controller exhibits superior tracking performance. Compared to the fuzzy PID controller, the LADRC controller demonstrates a reduction in the error magnitude for real-time dynamic tillage depth control. The LADRC controller demonstrates excellent real-time depth control performance, even when operating under uneven road conditions and in the presence of other disturbances.

Randomized Road Disturbance Signal of Unevenness.

Performance of real-time control of different tillage depths under disturbing signals.

The fuzzy PID control and LADRC control systems were employed to regulate the rotary tilling depth of the electric rotary tiller based on varying operational conditions. The detailed parameter settings of the fuzzy PID controller and LADRC controller are shown in Table 3. The setting of fuzzy PID parameters is an iterative and experimental process that may require several adjustments and simulation tests to achieve optimal control performance. In contrast, good control results can be achieved with the LADRC controller by setting only three simple parameters. The efficacy of the two control strategies in automatic tillage depth control systems was subsequently assessed. Furthermore, the quality of rotary tillage under both control modes was assessed to determine whether the requisite standards of uniformity and stability for tillage depth control are satisfied25,26.

As illustrated in Fig. 13, this electric rotary tiller is primarily utilized for intercropping between tea trees and soybeans in tea plantations. The length of the test tea plantation is 20 m, and the average rotary speed of the rotary tiller in the experiment is set to 150 r/min27. Considering the lightweight and low-power design of the electric rotary tiller, it is possible for the rotary speed to exceed the optimal range, potentially resulting in slippage of the tiller. Additionally, excessive tillage depth may also contribute to this phenomenon. Incomplete tillage operations and other abnormal working conditions may also occur. Through testing of the electric rotary tiller in the experimental field, it was found that the travel speed should not be too high, and the speed range should be controlled between 0.5 and 1.0 km/h. Accordingly, two average operating speeds were established: 0.5 km/h and 0.8 km/h. The recommended planting depth for soybeans is 50 mm, with an optimal soil tillage depth ranging from 80 to 100 mm. Consequently, preset tillage depths of 80 mm and 100 mm were selected for each operating speed. The combinations of average operating speeds and tillage depths are presented in Table 4, which delineates four distinct groups.

Electric Tiller Working Experiment.

The tillage depth adjustment system was tested under the four operating conditions described above. Furthermore, the rotary tiller was considered to be operating stably within a 20 m distance, with an additional 5 m on either side required to achieve the desired tillage depth. Each experiment was repeated three times under different operational conditions, with the average value recorded as the final result. The electric rotary tiller was tested using two control systems: LADRC and fuzzy PID. The tests were conducted independently as follows:

The machine status was verified to ensure that the control system, walking system, and rotary tiller system were operating correctly.

Either the fuzzy PID control algorithm or the LADRC control algorithm was configured.

The timer was started, and the electric rotary tiller began operation, with the tiller unit adjusted to achieve the target depth.

Sampling points were marked at 25 cm intervals along the rotary tiller’s path, selecting a total of 40 sampling points.

Once the electric rotary tiller completed its operation, the actual tillage depth was manually measured at the marked points.

The aforementioned methods were employed to execute the experiments, the results of which are presented in Fig. 14. This figure illustrates the tillage depth values at each marked point under varying working conditions. After the data were organized, the tillage depth curves for the different working conditions are presented in Fig. 15.

Data collection on the depth of rotary tillage.

Rotary tillage depth under different control methods.

Table 5 provides a quantitative evaluation of the errors for the two control algorithms. It shows the data for the maximum absolute error, mean absolute error, and standard deviation of the absolute error for each algorithm under different operating conditions. The maximum absolute error of the proposed LADRC controller is 10.1 mm at different working speeds, which is significantly lower than that of the fuzzy PID controller (20.1 mm). In terms of mean absolute error, the LADRC controller reduced the error by 57.7%, 66.3%, 56.1%, and 42.1% in comparison with the fuzzy PID controller across the four operating conditions. Furthermore, the standard deviation of the absolute error was reduced by 63.2%, 63.4%, 58.3%, and 37.8%, respectively. The above analysis indicates that the mean absolute error should be controlled within 10% when following the desired tillage depth. The proposed algorithm outperforms the other algorithms in terms of both the mean tracking error and its variability. Typically, the faster the working speed of the electric rotary tiller, the greater the error is expected to be. Due to the lack of prior field preparation and the random distribution of soil hardness in the actual low-speed experiments, the rotary tilling test was first conducted at 0.5 km/h, leading to the observation that the error at 0.5 km/h was greater than that at 0.8 km/h. Overall, the proposed algorithm maintains good tracking performance and high robustness.

The experimental results demonstrated that fluctuations throughout the entire rotary tilling process remained within a stable range. The results obtained in the experiments varied depending on the specific working conditions, including preset tillage depths and operating speeds. As shown in Fig. 15; Table 6, the electric rotary tiller, utilising both fuzzy PID and LADRC control strategies, quickly attained the desired tillage depth during actual operations. As illustrated in Fig. 15, the maximum discrepancy between the actual and preset tillage depths across diverse operational conditions was 23 mm, while the minimum was 16 mm. The mean standard deviation of the tillage depth was 10.5 mm, and the mean coefficient of variation in tillage depth stability was 11.9%. The maximum deviation between the actual and preset tillage depths under different working conditions was 8 mm when the LADRC control method was used, with a minimum deviation of 6 mm. The mean standard deviation of the tillage depth was 3.2 mm, with a mean coefficient of variation in tillage depth stability of 3.7%. Compared to the fuzzy PID control method, the LADRC control strategy demonstrated a reduction in the variation in tillage depth stability by 68.9%. Furthermore, the rotary tiller equipment utilizing ADRC control noticeably reduced the fluctuations in tillage depth, while maintaining stability. This resulted in a reduction in the variation rate of tillage depth and an improvement in the quality of rotary tillage operations.

This paper proposes a novel design method for a tillage depth control system based on body posture sensors, grating sensors, control units, and hybrid stepper motors, with the aim of addressing the issue of uneven tillage depth control in electric rotary tillers. The STM32 microcontroller serves as the control unit, with the LADRC control strategy applied innovatively to the tillage depth control system. The system achieves a control response time of 0.14 s, ensuring high precision, rapid response speed, and minimal fluctuation in tillage depth control. The electric rotary tiller utilizing the LADRC control strategy is capable of maintaining stability in tillage depth throughout the operational cycle. In comparison to the fuzzy PID control, the LADRC strategy demonstrated a reduction in the variation in tillage depth stability by 68.9%, thereby meeting the consistency requirements for tillage depth. The system achieves a fully adaptive closed-loop control process for tillage depth by collecting real-time signals from the grating sensor and comparing the actual tillage depth with the target value. The proposed control system is capable of achieving automated tillage depth regulation, maintaining consistent tillage depth, improving soil quality, reducing the operator’s labor intensity, and enhancing the intercropping production efficiency and tillage quality in tea gardens.

All data generated or analysed during this study are included in this published article [and its supplementary information files].

Liu, W., Yang, R., Li, L., Zhao, C. & Li, G. Energy and environmental evaluation and comparison of a diesel-electric hybrid tractor, a conventional tractor, and a hillside mini-tiller using the life cycle assessment method. J. Clean. Prod. 469, 143232 (2024).

Article CAS Google Scholar

Sahoo, A. U. & Raheman, H. Development of an electric reaper: a clean harvesting machine for cereal crops. Clean. Technol. Environ. Policy. 22, 955–964 (2020).

Article MATH Google Scholar

Fawzi, H. et al. TOQO: a new Tillage Operations Quality Optimization model based on parallel and dynamic decision support system. J. Clean. Prod. 316, 128263 (2021).

Article MATH Google Scholar

Kim, Y-S. et al. Analysis of the effect of tillage depth on the working performance of tractor-moldboard plow system under various field environments. Sensors 22 (7), 2750 (2022).

Article ADS MathSciNet PubMed PubMed Central MATH Google Scholar

Wang, A. et al. Tillage depth regulation system via depth measurement feedback and composite sliding mode control: a field comparison validation study. Meas. Control. 57 (6), 685–702 (2024).

Article MATH Google Scholar

Liu, C. et al. Benefits of mechanical weeding for weed control, rice growth characteristics and yield in paddy fields. Field Crops Res. 293, 108852 (2023).

Article MATH Google Scholar

Liu, G. et al. Research on an intelligent vibration detachment system for rotary tiller based on soil surface roughness dynamic characteristics. Computers Electron. Agric. 224, 109214 (2024).

Article MATH Google Scholar

Xiao, M. et al. Design and experiment of fuzzy-PID based tillage depth control system for a self-propelled electric tiller. Int. J. Agricultural Biol. Eng. 16 (4), 116–125 (2023).

Article MATH Google Scholar

Hu, K., Zhang, W., Qi, B. & Ji, Y. Tillage depth dynamic monitoring and precise control system. Measurement and Control. 0(0):00202940241263454 (2024).

Wang, Q., Wang, X., Wang, W., Song, Y. & Cui, Y. Joint control method based on speed and slip rate switching in plowing operation of wheeled electric tractor equipped with sliding battery pack. Comput. Electron. Agric. 215, 108426 (2023).

Article MATH Google Scholar

Kim, Y-S. et al. Development of a real-time tillage depth measurement system for agricultural tractors: application to the effect analysis of tillage depth on draft force during plow tillage. Sensors 20 (3), 912 (2020).

Article ADS MathSciNet PubMed PubMed Central MATH Google Scholar

Hu, K., Zhang, W., Qi, B. & Ji, Y. J. M. Control. Tillage depth dynamic monitoring and precise control system. :00202940241263454. (2024).

Jia, H. et al. An adaptable tillage depth monitoring system for tillage machine. Biosyst. Eng. 151, 187–199 (2016).

Article MATH Google Scholar

Luo, C. et al. Research on the slip rate control of a power shift tractor based on wheel speed and tillage depth adjustment. Agronomy 13 (2), 281 (2023).

Article MATH Google Scholar

Sabouri, Y. et al. Development and laboratory evaluation of an online controlling algorithm for precision tillage. Sensors 21 (16), 5603 (2021).

Article ADS PubMed PubMed Central MATH Google Scholar

Li, H., Hu, X. & Cui, L. Magnetic field analysis for the permanent magnet spherical motor with SMC core. IEEE Trans. Magn. 59 (6), 1–9 (2023).

Article MATH Google Scholar

Groenhuis, V., Rolff, G., Bosman, K., Abelmann, L. & Stramigioli, S. Multi-axis electric stepper motor. IEEE Rob. Autom. Lett. 6 (4), 7201–7208 (2021).

Article Google Scholar

Wu, S., Shi, T., Guo, L., Wang, H. & Xia, C. Accurate analytical method for magnetic field calculation of interior PM motors. IEEE Trans. Energy Convers. 36 (1), 325–337 (2020).

Article ADS MATH Google Scholar

Hojati, M., Baktash, A. & Mukhopadhyay, S. C. Investigation of torque and sensitivity analysis of a two-phase hybrid stepper motor. Int. J. Smart Sens. Intell. Syst. ;17(1), 1-21 (2024).

Han, J-Q. Nonlinear design methods for control systems. IFAC Proceedings Volumes. ;32(2):1531-6. (1999).

Han, J. From PID to active disturbance rejection control. IEEE Trans. Industr. Electron. 56 (3), 900–906 (2009).

Article MATH Google Scholar

Gao, Z., Yan, H. & Zhou, X. (eds) Active Disturbance Rejection Control of Microgrid DC-DC Converter Based on AC Algorithm. 2024 IEEE International Conference on Mechatronics and Automation (ICMA); : IEEE. (2024).

Gao, Z. (ed) Editor Scaling and bandwidth-parameterization Based Controller Tuning (Acc, 2003).

MATH Google Scholar

Jin, H., Song, J., Lan, W. & Gao, Z. On the characteristics of ADRC: a PID interpretation. Sci. China Inform. Sci. 63 (10), 209201 (2020).

Article MATH Google Scholar

Kim, Y-S. et al. DEM simulation for draft force prediction of moldboard plow according to the tillage depth in cohesive soil. Computers Electron. Agric. 189, 106368 (2021).

Article ADS MATH Google Scholar

Liu, K., Sozzi, M., Gasparini, F., Marinello, F. & Sartori, L. Combining simulations and field experiments: effects of subsoiling angle and tillage depth on soil structure and energy requirements. Computers Electron. Agric. 214, 108323 (2023).

Article Google Scholar

Chen, P. et al. Effect of varying remote cylinder speeds on plough-breast performances in alternative shifting tillage. Computers Electron. Agric. 181, 105963 (2021).

Article MATH Google Scholar

Download references

This research was funded by the Natural Science Foundation of Fujian Province (grant nos. 2022J011191 and 2024J01909), and the Nanping Science and Technology Plan Project (grant nos. N2023Z001, N2023Z002, N2023J001 and NP2021KTS07).

The Key Laboratory for Agricultural Machinery Intelligent Control and Manufacturing of Fujian Education Institutions, Wuyi University, Nanping, 354300, Fujian, China

Wei Tao, Bin Chen, Xinkun Yang, Bo Guo, Wanwan Xu, Shaoye Ke & Shenghong Huang

Fujian Key Laboratory of Big Data Application and Intellectualization for Tea Industry, Wuyi University, Nanping, 354300, Fujian, China

Wei Tao & Bin Chen

You can also search for this author in PubMed Google Scholar

Conceptualization, W.T. and B.C.; methodology, W.T.; software, W.X.; validation, X. Y., B. G., B.C. and S.K; data curation, S. H.; writing—original draft preparation, B.C.; writing—review and editing, W.T.

Correspondence to Bin Chen.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Below is the link to the electronic supplementary material.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Tao, W., Chen, B., Yang, X. et al. Design and experimental study of tillage depth control system for electric rotary tiller based on LADRC. Sci Rep 15, 1740 (2025). https://doi.org/10.1038/s41598-025-86283-6

Download citation

Received: 31 October 2024

Accepted: 09 January 2025

Published: 11 January 2025

DOI: https://doi.org/10.1038/s41598-025-86283-6

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative