Direct-drive power system of low-altitude rotary cultivator and control method of direct-drive power system

Abstract

The invention relates to the technical field of tillage equipment, and discloses a direct-drive power system of a low-altitude rotary cultivator and a control method of the direct-drive power system. A data acquisition module of the system acquires environment and operation data in the operation process of the low-altitude rotary cultivator; the expected cultivation parameter generation module builds a cultivation adaptive adjustment model, and inputs the collected data into the model to obtain expected cultivation parameters; the initial-stage control signal module generates an initial-stage control signal of the direct-drive power system based on the expected cultivation parameters; the initial execution module executes an initial tillage action according to the initial control signal; the deviation collection module collects actual cultivation data after the initial cultivation action and calculates the deviation between the actual cultivation data and the expected cultivation parameters; the control correction module corrects the initial control signal based on the deviation and generates a final control signal. The system adapts to a complex farming environment through self-adaptive adjustment and dynamic correction, the farming quality and the equipment stability are improved, and the precise farming requirement is met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned aerial farming equipment, in particular to a direct-drive power system of a low-altitude rotary cultivator and a control method thereof. BACKGROUND

[0002] In the development process of agricultural mechanization, the existing unmanned rotary cultivators are gradually applied to the cultivation of small plots of land and complex terrain areas due to their flexible operation characteristics, and become an important supplement to traditional large-scale cultivation machinery.

[0003] Currently, the power systems of unmanned rotary cultivators on the market mostly adopt indirect transmission structures, which transmit power to the cultivation execution mechanism through transmission components such as gears and belts. This transmission method not only has the problem of power loss, but also easily leads to a decrease in operation stability due to wear of the transmission components, increasing equipment maintenance costs and downtime for repair. The control method of the existing unmanned rotary cultivators is mostly a preset fixed parameter mode, that is, after setting the cultivation speed, depth and other parameters according to the experience of the operator, the power system operates according to the fixed parameters. However, there are many variables in the actual cultivation environment, such as differences in soil moisture in different plots or even different areas of the same plot, and the cultivation resistance of dry soil and wet soil is completely different; the soil texture is also divided into clay, loam and sand, and the hardness and looseness of each type of soil are significantly different; in addition, factors such as the density of weeds and crop residues in the field, differences in plot topography, scattered distribution of plots, and sudden changes in local soil conditions, etc. can also directly affect the cultivation process.

[0004] Under the influence of these complex and variable environmental factors, the fixed parameter control method is difficult to meet the actual needs. When encountering areas with high soil resistance, fixed cultivation parameters may cause the power system to be overloaded, affecting the cultivation efficiency and possibly causing damage to the power components; while in areas with low soil resistance, fixed parameters may result in shallow cultivation depth or excessive cultivation speed, failing to achieve the ideal cultivation effect, and thus affecting subsequent seeding and crop growth. Therefore, there is a lack of dynamic adjustment mechanism for the tillage process in the prior art, even if some devices have data acquisition function, the collected environmental and operation data cannot be effectively converted into the basis for adjusting the operation parameters of the dynamic system, so that the deviation occurring in the tillage process cannot be corrected in time, and the overall tillage quality is uneven, which is difficult to meet the requirements of modern agriculture for precision tillage. In the current agricultural production field, unmanned rotary tillers have gradually replaced traditional manual operation devices, and the tillage efficiency is improved by means of automatic control, but it is still limited by the ground driving conditions, and the operation flexibility is insufficient in scattered farmland and complex terrain areas; and the low-altitude weeding machine relies on a low-altitude mobile platform, and can break through the ground driving limit, flexibly adapt to complex scenes such as small plots of land, transitional farmland around forest land, and efficiently complete the weeding operation, and the technical characteristics of low-altitude movement and dynamic operation provide a new idea for the upgrading of tillage equipment. On this basis, the low-altitude rotary tiller, as a new type of equipment suitable for complex farmland scenes, can carry out operation in the areas that the traditional rotary tiller cannot cover, further expanding the utilization range of farmland, and also puts forward new requirements for the dynamic management of farmland resources - the quantity, quality and spatial distribution changes of the farmland resources involved in the operation of the low-altitude rotary tiller need to be accurately calculated to ensure the scientificity and accuracy of the replacement of such farmland resources and forest land resources. SUMMARY

[0005] The purpose of the present application is to provide a direct-drive power system of a low-altitude rotary tiller and a control method thereof, which dynamically adjusts the comprehensive control signal of the low-altitude rotary tiller by intelligently collecting environmental and operation data in the flight and operation process, so as to solve the problems raised in the background art.

[0006] To achieve the above-mentioned purpose, the present application provides a direct-drive power system of a low-altitude rotary tiller, which comprises: a data acquisition module for acquiring environmental and operation data during the operation of the low-altitude rotary tiller; an expected tillage parameter generation module for building a tillage self-adaptive adjustment model and inputting the environmental and operation data to obtain expected tillage parameters; an initial control signal module for generating an initial control signal of the direct-drive power system based on the expected tillage parameters; an initial execution module for executing an initial tillage action based on the initial control signal; a deviation collection module for collecting actual tillage data after the initial tillage action and calculating the deviation from the expected tillage parameters; a control correction module for correcting the initial control signal based on the deviation to generate a final control signal; The low-altitude rotary tiller comprises a rotary tiller carried by a drone or a rotary tiller of a unmanned helicopter.

[0007] Preferably, the data acquisition module comprises: The sensor is arranged on the low-altitude rotary cultivator to collect soil characteristic data and position data; The soil characteristic data and the position data are analyzed to obtain a cultivation condition parameter of the operation area; The environment and operation data are extracted based on the cultivation condition parameter.

[0008] Preferably, the expected cultivation parameter generation module comprises: The soil structure of the operation area is analyzed to obtain a cultivation self-adaptive parameter, and a cultivation self-adaptive adjustment model is built; The environment and operation data are input into the cultivation self-adaptive adjustment model, and an expected cultivation depth and an expected cultivation speed are obtained by analyzing an energy conversion relationship; The expected cultivation parameter is generated by combining the expected cultivation depth and the expected cultivation speed.

[0009] Preferably, the initial control signal module comprises: An initial cultivation expected value is generated according to a difference between a real-time cultivation value of the operation area and the expected cultivation parameter; An initial basic control parameter of the direct-drive power system is generated based on the initial cultivation expected value; An initial control signal of the direct-drive power system is generated through a control parameter analysis algorithm.

[0010] Preferably, the initial execution module comprises: The direct-drive power system performs an initial cultivation action based on the initial control signal; After the initial cultivation action is completed, an actual cultivation depth and an actual cultivation speed are recorded; The actual cultivation depth and the actual cultivation speed are transmitted to the deviation collection module.

[0011] Preferably, the deviation collection module comprises: The actual cultivation data are received and compared with the expected cultivation parameter; A depth deviation between the actual cultivation depth and the expected cultivation depth is calculated; A speed deviation between the actual cultivation speed and the expected cultivation speed is calculated; The depth deviation and the speed deviation are integrated to generate a cultivation deviation.

[0012] Preferably, the control correction module comprises: A control error is analyzed based on the cultivation deviation; A basic control parameter correction value of the direct-drive power system is generated through an adaptive control algorithm; A final basic control parameter is generated by correcting the initial basic control parameter based on the basic control parameter correction value; The end-stage basic control parameter is analyzed by a control parameter analysis algorithm to generate an end-stage control signal.

[0013] Preferably, the system further comprises an end-stage execution module for executing an end-stage tillage action based on the end-stage control signal. The end-stage execution module comprises: The direct-drive power system executes the end-stage tillage action based on the end-stage control signal. After the end-stage tillage action is completed, end-stage tillage data is collected. The tillage efficiency of the end-stage tillage action is evaluated.

[0014] Preferably, the system further comprises a safety warning module for monitoring system state parameters and performing safety warning. The safety warning module comprises: The temperature parameter and the pressure parameter of the direct-drive power system are monitored. An alarm is issued when the temperature parameter exceeds a temperature safety threshold. An alarm is issued when the pressure parameter exceeds a pressure safety threshold. The operating state of the direct-drive power system is adjusted.

[0015] Preferably, the application further comprises a direct-drive power control method for a low-altitude rotary tiller, applied to the direct-drive power system of the low-altitude rotary tiller, comprising: Environmental and operation data are collected. Expected tillage parameters are obtained based on the environmental and operation data. An initial control signal of the direct-drive power system is generated based on the expected tillage parameters. An initial tillage action is executed based on the initial control signal. Actual tillage data is collected and a deviation is calculated. The initial control signal is corrected based on the deviation to generate an end-stage control signal. An end-stage tillage action is executed based on the end-stage control signal. System state parameters are monitored for safety warning.

[0016] Compared with the prior art, the application has the following beneficial effects: 1. The direct-drive power system of the low-altitude rotary tiller first discards the traditional indirect transmission structure and adopts a direct-drive mode to drive the tillage execution mechanism, reducing the use of intermediate transmission components such as gears and belts, reducing the power loss in the transmission process from the structure, avoiding the failure problems caused by the wear of intermediate transmission components, reducing the frequency and cost of equipment maintenance, prolonging the overall service life of the power system, improving the stability and reliability of equipment operation, and ensuring that the tillage operation can be continuously and efficiently performed.

[0017] 2. In terms of control logic, the system captures environmental and operational data in real time during the tillage process through a data acquisition module. This allows for a comprehensive understanding of key information such as soil moisture, texture, weed distribution, equipment operating speed, and current tillage depth, providing comprehensive and accurate basic data for subsequent parameter adjustments and breaking the limitations of traditional fixed-parameter control modes. The adaptive tillage adjustment model built by the expected tillage parameter generation module integrates and analyzes the collected multi-dimensional data, automatically generating suitable expected tillage parameters based on different environmental conditions. This ensures that the operating parameters of the power system match the actual tillage environment, avoiding poor tillage results or equipment overload caused by environmental changes.

[0018] 3. The initial control signal module generates initial control signals based on the expected tillage parameters, guiding the initial execution module to complete the initial tillage actions. This ensures that the tillage operation can start according to the parameters adapted to the environment, laying a good foundation for subsequent stable tillage. The deviation collection module, after the initial tillage actions, promptly collects actual tillage data and calculates the deviation from the expected tillage parameters. This allows for precise detection of differences between the actual operation and the ideal state, such as whether the tillage depth meets expectations or whether the tillage speed is suitable for the current soil resistance. 4. The control correction module corrects the initial control signal based on the aforementioned deviations, generating the final control signal and achieving dynamic adjustment during the tillage process. This dynamic correction mechanism can respond in real time to changes in the tillage environment. Even in the same plot where soil conditions change abruptly, the system can quickly adjust the operating parameters of the power system to keep the tillage effect within the ideal range. For example, when entering an area with high soil resistance, the system can appropriately reduce the tillage speed or adjust the power output through the correction control signal to avoid overloading the power system; when entering an area with low soil resistance, it can appropriately increase the tillage efficiency while ensuring that the tillage depth meets the standards. Attached Figure Description

[0019] Figure 1 This is a schematic diagram illustrating the working principle of the direct-drive power system of the unmanned aerial vehicle-borne rotary tiller in this embodiment. Figure 2 This is a flowchart of the data acquisition module; Figure 3 This is a flowchart of the initial control signal module; Figure 4 This is a flowchart of the deviation collection module. Detailed Implementation

[0020] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0021] The low-altitude rotary cultivator described in the present application refers to unmanned equipment for low-altitude cultivation operation, which covers unmanned aerial rotary cultivators and unmanned helicopter rotary cultivators, and can be applied to cultivation operation in small plots of land and complex terrain areas, as a supplement to traditional large-scale cultivation machinery, and the core requirement is to achieve precise cultivation through efficient power transmission and flexible parameter adjustment, which is completely suitable for the design goal of the direct-drive power system of the present application.

[0022] The low-altitude rotary cultivator in the present application can be an unmanned aerial rotary cultivator or an unmanned helicopter rotary cultivator. Rotary cultivators that can obtain environmental data such as geographic location, soil characteristics, and weed conditions during flight should be considered as low-altitude rotary cultivators in the present application.

[0023] First embodiment The unmanned helicopter rotary cultivator is similar in form to a large-scale plant protection helicopter, with a rotary cultivator suspended below. The unmanned aerial rotary cultivator is the most common form, which uses a multi-rotor unmanned aerial vehicle as a flight platform and a power source, and a multi-rotor agricultural unmanned aerial vehicle as a flight platform. By mounting a dedicated rotary cultivation module, it is an intelligent equipment for aerial rotary cultivation.

[0024] The present embodiment schematically provides a direct-drive power system of an unmanned aerial rotary cultivator and a control method thereof, please refer to Figure 1 , the system comprises a data acquisition module, an expected cultivation parameter generation module, an initial control signal module, an initial execution module, a deviation collection module and a control correction module. The specific implementation is as follows: The data acquisition module is responsible for collecting environmental and operating data during the operation of the unmanned plane cultivator. These data include soil properties, geographic location, weed conditions, and real-time working condition information. The expected tillage parameter generation module constructs a tillage adaptive adjustment model. The collected data are input into the model, and the expected tillage depth and expected tillage speed are output through internal algorithm processing, which together constitute the expected tillage parameters. The initial control signal module generates an initial tillage expected value according to the difference between the expected tillage parameters and the real-time tillage values, and then derives the initial basic control parameters of the direct-drive power system, and converts them into specific initial control signals through a control parameter analysis algorithm. The initial execution module receives the signals and drives the direct-drive power system to perform the initial tillage action, and records the actual tillage depth and actual tillage speed after completion. The deviation collection module compares the actual data with the expected parameters, calculates the depth deviation and speed deviation, and integrates them into the tillage deviation. The control correction module analyzes the control error based on the deviation, generates a basic control parameter correction value using an adaptive control algorithm, modifies the initial basic control parameters, obtains the final basic control parameters, and finally generates the final control signal through the control parameter analysis algorithm. The system realizes dynamic adjustment of the tillage process through a closed-loop feedback mechanism, improving the operation accuracy and adaptability.

[0025] Example 1: see Figure 2In the actual operation of the unmanned rotary cultivator, the deployment of the data acquisition module is the starting point of the system operation. The core of this module is to install a multi-type sensor array at the key parts of the rotary cultivator. Specifically, a soil compaction sensor is arranged near the installation base of the rotary cultivator blade shaft. This sensor directly contacts the soil through an extendable probe and measures the resistance of the soil to the probe, thereby converting the physical hardness of the soil into an electrical signal. Inside the unmanned aerial vehicle landing gear, a soil moisture sensor is installed. This sensor uses the frequency domain reflection principle to infer the dielectric constant of the soil by emitting high-frequency electromagnetic waves and analyzing their returned waveforms, and then calculates the volumetric water content. The unmanned rotary cultivator uses a multi-rotor unmanned aerial vehicle as the flight platform, and hangs a special rotary cultivation module below. At the bottom of the rotary cultivation module, a high-precision real-time kinematic GPS receiver and an image acquisition device are fixed beside the rotary cultivator blade shaft. The bottom of the rotary cultivator is equipped with a camera all-in-one machine core, which is a highly integrated camera equipment that combines lens, image sensor, image processing circuit, and other functional modules together to form a compact and efficient camera system. The GPS receiver can receive satellite signals and correct them with data from the ground reference station, and can output the latitude and longitude coordinates and elevation information of the rotary cultivation module in real time, with a positioning accuracy of centimeters. The image acquisition device faces the area to be cultivated and synchronously acquires image information of the working area. After data fusion, the weed density and crop residue distribution characteristics of the area to be cultivated can be accurately identified, and these feature information can be associated with specific geographic locations, providing accurate environmental data support for subsequent cultivation parameter adjustment.

[0026] These sensors work continuously during the cultivation process, and the sampling frequency of environmental data, soil property data, and location data is collected and transmitted through the onboard data bus of the unmanned aerial vehicle.

[0027] The raw data collected needs to be processed in a series of steps before it can be turned into useful information. The raw voltage or current signal is first converted into a digital signal by an analog-to-digital converter. Subsequently, the embedded processor filters the digital signal to remove high frequency noise caused by mechanical vibration or electromagnetic interference. The cleaned data after pre-processing is sent to a data fusion algorithm unit. This unit aligns and synchronizes data from different sources by time stamp. For example, the precise geographic location measured by GPS at a certain time is bound to the readings of the soil moisture sensor and the firmness sensor at the same time. Through the aligned data set, the system begins to analyze the farming condition parameters of the operating area. The system has a pre-stored database of characteristics of different soil types. By pattern matching the real-time measured soil moisture and firmness data with the database, it can be determined that the soil type of the current operating area belongs to clay, loam or sand. At the same time, combined with the elevation data provided by GPS, the processor can obtain the slope information of the land by calculating the elevation change between consecutive position points. All these information: soil type, average moisture level, hardness distribution and slope together constitute the farming condition parameters of the operating area. The environmental and operating data are the key quantitative indicators extracted from these condition parameters, for example, the system will record the average soil firmness value of the current area as an important operating data, or take the maximum slope value as a key environmental data that affects the flight attitude of the unmanned aerial vehicle. These extracted data are structured into data packets, waiting to be transmitted to the next module.

[0028] After the data acquisition module obtains the image information of the work area and GPS positioning data, for the calculation of weed density and crop residue distribution characteristics, the image segmentation algorithm is first used to distinguish the weed, crop residue and soil area in the image, and then the pixel statistics are used to determine the proportion of weeds and residues in the image, combined with the actual plot area associated with GPS, to convert the weed density at the corresponding position; At the same time, according to the difference of image gray value, the weed class (such as the difference of rhizome hardness) and the crop residue height are identified, and the residue distribution characteristics are clarified. After the calculation is completed, these data are transmitted to the expected tillage parameter generation module, which is combined with the weed density and residue distribution characteristics in the built tillage adaptive adjustment model to adjust the expected parameters - if the weed density in a certain area is high or the rhizome is relatively hard, the expected tillage power output parameter is increased; If the crop residue is high, the expected tillage depth parameter is increased. The initial control signal module generates adaptive initial control signals based on the adjusted expected parameters, such as increasing the direct drive motor torque setting for weed dense areas and issuing the rotary tillage blade shaft depth adjustment instruction for high residue areas to drive the rotary tillage module to perform corresponding tillage actions; When the subsequent deviation collection module detects that the weeds are not completely removed or the residues are not sufficiently crushed, it will include this situation in the tillage deviation calculation, and the control correction module will correct the control signal accordingly to further optimize the power output and tillage depth of the rotary tiller, realizing the cooperative adaptation of the unmanned aerial vehicle flight platform and the rotary tillage module, and ensuring the tillage effect.

[0029] After the expected tillage parameter generation module receives the structured data packet from the data acquisition module, it starts its analysis process. This module first focuses on analyzing the soil structure of the operating area. Soil structure analysis is a deep information extraction process that not only identifies soil types, but also infers the hierarchical structure and mechanical properties of the soil. The system will integrate humidity, compactness, and historical tillage data (if available) to assess the soil's adhesion, cohesion, and shear strength. For example, clay in high humidity may be judged to have high adhesion and plasticity, while dry sand is judged to be loose and low in cohesion. Based on these analyses, a set of adaptive tillage parameters is determined. These parameters include the friction coefficient of the soil and the rotary tillage blade, the maximum resistance the soil may generate during tillage, and a predictive model of the soil's response at different depths. These adaptive parameters are the basis for building the tillage adaptive adjustment model.

[0030] The tillage adaptive regulation model is essentially a rule-based and physics-based decision system. It treats the soil as a dynamic reactor, and the core of the model is to simulate the energy conversion process of the interaction between the tillage blade and the soil. The model calculates the energy required for the tillage blade to cut the soil, which includes the work done to cut off the grass roots, break up the soil clumps, and overcome the soil friction. At the same time, the model also continuously receives the available energy data from the UAV power system, i.e. the maximum torque and power that the motor can currently output. Environmental and operational data such as soil moisture, compactness, and slope are injected as key input variables into this energy balance model. High soil compactness means more energy is needed to break up the soil, and a larger slope requires the UAV to allocate more power to maintain attitude stability, which may reduce the power used for tillage. The model iteratively calculates to find the optimal solution that can effectively complete the tillage task under the current environmental constraints without exceeding the power system load. The output of this optimal solution is two key values: the expected tillage depth and the expected tillage speed. The expected tillage depth is a result of balancing between power constraints and agronomic requirements, and the expected tillage speed is to ensure that the tillage area per unit time and power consumption are balanced at a given depth.

[0031] Example 2: refer to Figure 3 After the UAV tillage machine enters the predetermined tillage area, the initial control signal module starts to work. The core task of this module is to generate precise control instructions to drive the direct-drive power system. The module first receives a data packet from the expected tillage parameter generation module, which contains the expected tillage parameters for the current field, including a target tillage depth of 15 cm and a target forward speed of 1.2 meters per second. At the same time, the ultrasonic depth sensor installed on the tillage machine blade shaft feeds back the current detected tillage depth in real time, for example 12 cm; the UAV flight control system calculates the actual forward speed through the airspeed meter and GPS data fusion, for example 1.0 meters per second. The system calculates the difference between the real-time tillage value and the expected tillage parameters: the tillage depth deviation is +3 cm (target 15 cm minus actual 12 cm), and the speed deviation is +0.2 meters per second (target 1.2 meters per second minus actual 1.0 meters per second). These differences are input into a difference processing unit.

[0032] Based on these differences, the system generates a primary tillage expectation value. This expectation value is a control vector containing two components: a depth compensation amount + 3 cm and a speed compensation amount + 0.2 m / s. This vector represents the immediate adjustments the system needs to make to reach the intended goal. The primary tillage expectation value is then sent to the control parameter generation unit. This unit contains a proportional-integral-derivative controller that calculates the corresponding power system adjustment parameters based on the input expectation value components. For the depth compensation amount, the controller outputs an instruction to increase the motor torque; for the speed compensation amount, the controller outputs an instruction to increase the motor speed. After calculation, the system determines the primary basic control parameters as follows: the target speed of the brushless DC motor is set to 850 rpm, the expected output torque is 35 Nm, and 70% of the total power of the power system is allocated to the rotary tillage operation, with the remaining 30% reserved for flight attitude maintenance.

[0033] The control parameter analysis algorithm receives these primary basic control parameters and begins signal conversion work. The algorithm accesses the pre-stored motor performance database, which records the working characteristic curves of specific motor models under different speeds and loads. For a target speed of 850 rpm, the algorithm queries the database to determine the corresponding optimal working voltage of 48 V and calculates the required PWM duty cycle of 65%. For an expected output torque of 35 Nm, the algorithm calculates the corresponding armature current reference value of 40 A based on the motor torque constant. The power distribution ratio of 70% is converted into specific power management instructions, limiting the maximum instantaneous power consumption of the rotary tillage system to no more than 2.5 kW. All these calculated parameters are integrated and encoded into a set of digital control instructions, which are sent to the direct-drive motor controller at a rate of 1000 frames per second through the controller area network bus.

[0034] The primary execution module receives and parses these digital control instructions. The core of the direct-drive power system is a high-power-density brushless DC motor whose output shaft is directly connected to the rotary tillage blade shaft through a flange without any intermediate transmission devices such as gears or belts. After receiving the 65% duty cycle PWM instruction, the motor controller drives its internal power field effect transistor to switch at a corresponding frequency, outputting a working voltage of 48 V to the motor winding. At the same time, the current control loop adjusts the output current based on the 40 A reference value to ensure that the motor generates an accurate 35 Nm torque. The rotary tillage blade shaft begins to rotate under the drive of this torque, and the blade cuts into the soil at a speed of 850 rpm. The UAV moves forward at an initial speed of 1.0 m / s under the control of the flight control system, performing the primary tillage action.

[0035] During the initial tillage action, multiple sensors continuously monitor the working status. A linear displacement sensor installed on the toolbar support arm measures the vertical position change of the toolbar relative to the UAV body in real time, with a measurement accuracy of millimeter level. When the toolbar is displaced due to soil resistance, the sensor detects the displacement amount through voltage change, which is converted into tillage depth value after calibration. At the same time, the main flight controller of the UAV comprehensively processes data from the global navigation satellite system receiver, inertial measurement unit and airspeed meter, and calculates the accurate ground speed through sensor fusion algorithm. When the system detects that the rotary tiller has completed a standard working distance, for example, 20 meters forward, the initial tillage stage is declared complete.

[0036] The full-range average tillage depth recorded by the displacement sensor is 13.5 cm, and the average forward speed calculated by the flight control system is 1.05 m / s. These data are packaged as data packets containing timestamps, position information and quality indicators, and transmitted to the deviation collection module through a wireless data link. The data packet also contains auxiliary information, such as the average motor operating current of 38 amperes and the winding temperature of 65 degrees Celsius during the recording period, which provides additional context for subsequent analysis. At this point, the execution and data collection process of the initial tillage action is complete, and the system is ready to enter the deviation analysis stage.

[0037] Example 3: refer to Figure 4 The deviation collection module starts working after the initial tillage action is completed, and the module receives the transmission data packets from the initial execution module through a dedicated data interface. The data packets are packaged in a structured format, containing a timestamp field to mark the UTC time of data collection, a position information field to record the GPS latitude and longitude coordinates, a core data field to store the actual tillage depth value and the actual tillage speed value , and an auxiliary field containing motor current, winding temperature and other working condition parameters. The system first parses and verifies the data packets, checks the data integrity flag, verifies the correctness of the cyclic redundancy check code, and confirms that there is no error before extracting the actual tillage depth and the actual tillage speed from the core data field. These measured data are stored in floating point format, with depth unit in centimeters and speed unit in meters per second, with precision retained to one and two decimal places respectively.

[0038] The module reads the target value data generated by the expected tillage parameter generation module from the shared memory area. This data set contains the expected tillage depth and the expected tillage speed , with the same units as the measured values. The system establishes a data mapping relationship, accurately aligns the measured values and expected values for the same geographic coordinate point through timestamp and position information double matching. For each sampling point, the basic deviation amount is calculated by performing arithmetic subtraction operation: depth deviation component , velocity deviation component . These base deviation values are stored with sign, positive value indicates measured value exceeds expectation, negative value indicates failure to meet expectation.

[0039] Base deviation values need to be normalized to adapt to different dimensions. The system calls preset reference parameters, including maximum allowed tillage depth (e.g. 30 cm) and maximum designed tillage velocity (e.g. 3 m / s). The dimensionless conversion is realized by the following formula:

[0040] wherein: denotes normalized depth deviation rate, denotes normalized velocity deviation rate, both in percentage. The conversion eliminates dimensional difference, making deviations of different categories comparable.

[0041] The weight distribution unit dynamically adjusts the influence factor according to the tillage task requirements. The system has built-in weight coefficient configuration table, correlating different soil types and operation modes. When the data acquisition module identifies the current area as clay, the depth precision weight is automatically increased to 0.7, and the velocity weight is decreased to 0.3; if it is sandy loam, the adjustment is , . The weight coefficients satisfy the normalization constraint . The current operation mode parameters (such as deep tillage mode or fast shallow tillage mode) are read through the system state register, further fine-tuning the weight distribution proportion.

[0042] The normalized deviations and weight coefficients are input into the deviation synthesis processor. The processor performs weighted synthesis operation: first, calculate the weighted depth deviation term , then calculate the weighted velocity deviation term . The results of the two terms are input into the comprehensive deviation generator, and through the vector synthesis algorithm, output the two-dimensional deviation vector and one-dimensional comprehensive deviation scalar . The comprehensive deviation scalar is calculated using Euclidean norm:

[0043] The scalar value comprehensively reflects the deviation degree of the overall tillage state, the value range is 0%~100%, the larger the value, the more serious the deviation from the expected state.

[0044] Data validity verification mechanism is applied throughout the whole process. System sets physical reasonable range check: when working depth deviation exceeds ± 50%, data review process is triggered; when speed deviation exceeds ± 75%, sensor calibration program is started. Timing continuity check module analyzes mutation of adjacent data points, if deviation change rate of three consecutive sampling points exceeds 20% / second, data smoothing filter processing is inserted. The final generated working deviation data set includes: original deviation record table (stores deviation of each sampling point 、 ), standardized deviation record table ( 、 ), weight distribution record table ( 、 value and adjustment basis), comprehensive deviation vector and scalar value . All data are stored in deviation database after adding quality flag, and input parameters are provided to control correction module through memory sharing.

[0045] Example 4: After receiving working deviation data set from deviation collection module, control correction module starts control error analysis program. The data set includes standardized depth deviation rate, standardized speed deviation rate, weight distribution coefficient and comprehensive deviation scalar value. System first establishes deviation-error mapping table to convert working deviation into control error type identification. When depth deviation rate exceeds positive threshold, it is identified as insufficient depth control error; when speed deviation rate continuously deviates negatively, it is identified as excessive travel power control error. System maintains an error classification database, which includes twelve predefined error types and their characteristic patterns.

[0046] Error analysis unit adopts multi-level judgment logic. Primary judgment is based on instantaneous deviation value of current sampling point, middle-level judgment analyzes deviation change trend of the last ten sampling points, and high-level judgment integrates historical deviation records of current working area. When depth deviation rate remains in positive interval for three consecutive sampling points and shows an upward trend, the system determines that there is a progressive depth control error; when speed deviation rate frequently oscillates in positive and negative intervals, it is determined that there is a stability control error. Each determination result corresponds to error code and severity level. Adaptive control algorithm calls corresponding parameter correction strategy according to error type. System has multiple control algorithm modes built-in, including incremental regulation mode, proportional regulation mode and mode switching regulation mode. For progressive depth control error, the algorithm selects incremental regulation mode, which generates progressive adjustment instruction based on deviation history data; for stability control error, proportional regulation mode is enabled, which uses direct proportional function of current deviation value for calculation. The core of the algorithm is a state machine structure, which switches different calculation paths according to error code.

[0047] The generation process of basic control parameter correction values adopts a step-by-step calculation method. Taking motor speed correction as an example, the system first reads the current speed set value, then calculates the basic adjustment amount according to the speed deviation rate and weight coefficient. This adjustment amount is processed through a rate limiter to prevent the adjustment amplitude from being too large at a time. At the same time, the system dynamically compensates the adjustment amount with reference to motor temperature, voltage and other working condition parameters. The final output speed correction value contains three elements: adjustment direction (positive or negative), adjustment amplitude and adjustment duration. The same process is applied to the correction calculation of torque control parameters and power distribution parameters. The correction application unit performs parameter replacement operations, and this unit maintains two sets of data: the current control parameter set and the corrected control parameter set. When receiving a correction instruction, the system gradually transitions the current parameters to the target parameters. The transition process uses a smooth interpolation algorithm to avoid system oscillation caused by parameter mutation. Each parameter update generates a version record, including timestamp, parameter type, old value, new value, adjustment reason and other metadata. The system is equipped with a correction effect prediction mechanism. Before applying the correction value, the system behavior after correction is simulated through a digital twin model. The prediction results include the expected deviation change curve, system stability indicators and energy consumption estimates. When the prediction results show that the correction may cause system instability, the correction parameter re-optimization process is automatically triggered.

[0048] The entire correction process generates detailed operation logs, recording the input data, processing logic and output results at each decision point. These data form closed-loop learning samples for optimizing subsequent correction decisions. Referring to Table 1, the main parameter change records during a control correction process are shown.

[0049] Table 1: Control correction parameter change record table.

[0050]

[0051] The correction verification unit starts the monitoring program after parameter application, and this unit compares the deviation data before and after correction to calculate the improvement index. When a single correction does not achieve the expected effect, the system starts a secondary correction process according to the residual deviation amount. Secondary correction uses a more refined adjustment strategy, with half the adjustment step size and double the monitoring frequency. The system also includes a correction rollback mechanism. When new instability phenomena or system alarms occur after correction are detected, the system automatically reverts to the last stable parameter set. The rollback decision is based on multi-factor evaluation, including current deviation state, system load conditions and safety monitoring data. Each rollback operation generates an analysis report, recording the rollback reason and subsequent optimization suggestions.

[0052] The final set of basic control parameters generated contains complete state description information, and each parameter is labeled with a confidence level reflecting the stability and reliability of the parameter. High-confidence parameters are directly output, and low-confidence parameters require additional monitoring instructions. The parameter set is sent to the control parameter analysis algorithm module through the secure transmission protocol and enters the final control signal generation phase. The entire correction process realizes a complete closed loop from deviation identification to parameter optimization, ensuring that the system can adapt to changing operating conditions.

[0053] Example 5: After receiving the final control signal generated by the control correction module, the final execution module starts the operation process. This signal contains a complete set of parameter instructions, specifically specifying the target operating state of the direct-drive power system. The motor controller parses the speed set value, torque reference value, and power distribution ratio parameters in the signal. The state machine inside the controller switches to the final execution mode, which uses an enhanced control algorithm that increases the sampling frequency of the current loop and speed loop to twice that of the regular mode. The power drive unit adjusts the switching timing of the field effect transistor based on the parsed instructions to output an accurate voltage waveform to the brushless motor. The motor rotor accelerates to the target speed under the action of the electromagnetic field, and the acceleration process uses an S-shaped curve planning to avoid mechanical impact. The rotary tiller shaft directly obtains the rotary kinetic energy output by the motor through a rigid coupling, and the blade enters the soil layer at the corrected cutting angle and rotation speed.

[0054] During the final tillage action execution, the system activates the high-density data collection program. The strain gauge sensors installed on the flange of the blade shaft record the force changes of the blade at a sampling rate of 200 times per second and send them in real time to the central processor through the wireless transmission module. The flight control system of the unmanned aerial vehicle synchronously increases the positioning data update frequency, increasing the output rate of the fusion positioning data of GPS and inertial navigation system from the regular 10Hz to 50Hz. When the rotary tiller completes the preset work path segment, such as tillage along the east-west direction for 30 meters, the system automatically determines that the final tillage stage is complete. At this time, the data collection unit starts the summary program to extract the final tillage data from each sensor node. The key data sets include: average tillage depth, speed stability coefficient, blade shaft torque fluctuation range, motor winding temperature change curve, and unit distance energy consumption statistics.

[0055] The evaluation program is automatically executed after the data collection is completed. The evaluation system adopts a multi-dimensional index system, in which the uniformity of plowing is analyzed by comparing the standard deviation of depth at different sampling points. The system divides the plowing area into a virtual grid of 1 meter x 1 meter, calculates the deviation of the actual plowing depth of each grid unit from the target depth, and generates a depth distribution heat map. The unit area energy consumption calculation integrates the battery discharge data and rotary plowing operation time recorded by the flight control system, and is accurate to the number of watt-hours consumed per square meter. The operation time efficiency index is obtained by comparing the ratio of the actual plowing speed to the theoretical maximum speed, while considering the proportion of non-plowing time such as turning and climbing. The evaluation results generate a structured report, including index values, regional distribution characteristics and abnormal point labeling.

[0056] The safety warning module runs independently in the background of the system and obtains the state parameters of the power system through a dedicated data channel. The temperature monitoring unit includes three groups of redundantly arranged infrared temperature sensors, which are aimed at the positions of the motor shell heat sink, the controller power module and the bearing seat, respectively. The sensors collect surface temperature data at a frequency of 5 times per second, and after processing by the thermocouple cold end compensation algorithm, they are converted into accurate temperature values. The pressure monitoring uses a pressure resistance sensor array installed on the key nodes of the hydraulic circuit (if the system uses hydraulic assistance) or the motor axial load support structure. All monitoring data are input into the real-time analysis engine, which maintains a dynamically updated safety threshold database.

[0057] The threshold comparator continuously compares the real-time data with the safety threshold. The temperature threshold is divided into three levels: the first warning threshold is set to 20 degrees Celsius below the maximum allowable temperature of the motor insulation material, the second warning threshold is 10 degrees Celsius below the maximum allowable temperature, and the third emergency threshold is the material limit temperature. The pressure threshold is also set in stages, with a baseline value of 80% of the designed maximum working pressure. When the temperature parameter exceeds the first warning threshold, the system displays an amber warning icon on the operation interface and sends a warning message with status code 2001 to the ground control station. If the temperature continues to rise and breaks through the second warning threshold, the audible and visual alarm is triggered and an alarm log with high priority is generated. The pressure parameter out of limit adopts a similar staged response mechanism.

[0058] The state adjustment mechanism is activated synchronously upon alarm triggering. For temperature overrun events, the system first attempts software regulation: reducing motor output power setpoint, limiting maximum allowed current, and boosting cooling fan speed to full speed mode. If temperature does not fall back to safety range within 60 seconds of adjustment, a gradient load shedding procedure is executed: reducing output torque by 10% every 10 seconds until temperature returns to normal. For pressure overrun events, the control algorithm automatically reduces tillage depth setpoint and activates hydraulic system pressure relief valve (if installed). When either temperature or pressure parameters breach the tertiary emergency threshold, the system immediately executes a protective shutdown: cutting off motor power output, activating electromagnetic brakes, and switching the drone to hover mode. All state adjustment operations are recorded in the safety event log, including timestamp, adjustment action, adjustment magnitude, and post-adjustment parameter change profile.

[0059] The system maintenance module automatically generates a maintenance recommendation report after each job, which analyzes historical operation data to identify potential risk points. For example, when detecting that the motor winding temperature has reached 90% of the warning threshold for three consecutive jobs, it recommends cleaning the heat dissipation channel; when the bearing pressure fluctuation coefficient exceeds twice the reference value, it prompts to check the mechanical coaxiality. The report also includes usage time statistics and life prediction of key components, calculated based on the pre-set component replacement cycle model. All maintenance data is transmitted to the cloud server through an encrypted link to form a device lifecycle management file.

[0060] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one entity or action from another, without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0061] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A direct-drive power system for a low-altitude rotary tiller, characterized in that, include: The data acquisition module is used to collect environmental and operational data during the operation of the low-altitude rotary tiller; The expected tillage parameter generation module is used to build an adaptive adjustment model for tillage and input the environmental and operational data to obtain the expected tillage parameters; An initial control signal module is used to generate an initial control signal for the direct-drive power system based on the expected tillage parameters. The initial execution module is used to execute initial tillage actions based on the initial control signal; The deviation collection module is used to collect actual tillage data after the initial tillage action and calculate the deviation from the expected tillage parameters; A control correction module is used to correct the initial control signal based on the deviation to generate a final control signal; The low-altitude rotary tiller includes unmanned aerial vehicle (UAV) rotary tillers or unmanned helicopter-type rotary tillers.

2. The direct-drive power system of the low-altitude rotary tiller as described in claim 1, characterized in that, The data acquisition module includes: Sensors are installed on low-altitude rotary tillers to collect soil characteristic data and location data; Analyze soil property data and location data to obtain tillage condition parameters for the operational area; Environmental and operational data are extracted based on the aforementioned tillage condition parameters.

3. The direct-drive power system of the low-altitude rotary tiller as described in claim 1, characterized in that, The expected tillage parameter generation module includes: Analyze the soil structure of the operating area, obtain tillage adaptive parameters, and build a tillage adaptive adjustment model; The environmental and operational data are input into the adaptive tillage adjustment model, and the expected tillage depth and expected tillage speed are obtained by analyzing the energy conversion relationship. Expected tillage parameters are generated by combining expected tillage depth and expected tillage speed.

4. The direct-drive power system of the low-altitude rotary tiller as described in claim 1, characterized in that, The initial control signal module includes: The initial tillage expectation value is generated based on the difference between the real-time tillage value and the expected tillage parameter in the operating area. The initial basic control parameters of the direct-drive power system are generated based on the initial tillage expectations. The initial control signal for the direct-drive power system is generated through a control parameter analysis algorithm.

5. The direct-drive power system of the low-altitude rotary tiller as described in claim 1, characterized in that, The initial execution module includes: The direct-drive power system executes the initial tillage actions based on the initial control signal; After the initial tillage actions are completed, record the actual tillage depth and actual tillage speed; The actual tillage depth and actual tillage speed are transmitted to the deviation collection module.

6. The direct-drive power system of the low-altitude rotary tiller as described in claim 1, characterized in that, The deviation collection module includes: Receive actual tillage data and compare it with expected tillage parameters; Calculate the depth deviation between the actual tillage depth and the expected tillage depth; Calculate the speed deviation between the actual tillage speed and the expected tillage speed; Integrate depth deviation and velocity deviation to generate tillage deviation.

7. The direct-drive power system of the low-altitude rotary tiller as described in claim 1, characterized in that, The control correction module includes: Error control based on tillage deviation analysis; The basic control parameter correction values ​​of the direct-drive power system are generated through an adaptive control algorithm; The initial basic control parameters are corrected based on the basic control parameter correction values ​​to generate the final basic control parameters. The terminal control signal is generated by analyzing the terminal basic control parameters through the control parameter analysis algorithm.

8. The direct-drive power system of the low-altitude rotary tiller as described in claim 1, characterized in that, The system further includes: a terminal execution module, used to execute terminal tillage actions based on the terminal control signal; The final execution module includes: The direct-drive power system executes the final tillage action based on the final control signal; Collect final tillage data after the final tillage operations are completed; The efficiency of the final tillage action is evaluated.

9. The direct-drive power system of the low-altitude rotary tiller as described in claim 1, characterized in that, The system also includes a security early warning module, used to monitor system status parameters and issue security early warnings; The security early warning module includes: Monitor the temperature and pressure parameters of the direct-drive power system; An alarm is triggered when the temperature parameter exceeds the safe temperature threshold. An alarm is triggered when the pressure parameter exceeds the pressure safety threshold. Adjust the operating status of the direct-drive power system.

10. A direct-drive power control method for a low-altitude rotary tiller, applied to the direct-drive power system of the low-altitude rotary tiller as described in any one of claims 1-9, characterized in that, include: Collect environmental and operational data; Based on the aforementioned environmental and operational data, the expected tillage parameters are obtained; The initial control signal for the direct-drive power system is generated based on the expected tillage parameters. The initial tillage action is performed based on the initial control signal; Collect actual farming data and calculate deviations; The initial control signal is corrected based on the deviation to generate the final control signal; The final tillage action is executed based on the final control signal; Monitor system status parameters to provide safety warnings.

Images