Intelligent prediction system for threshing and separating quality of soybean combine

Abstract

The application discloses a soybean combine harvester threshing and separating quality intelligent prediction system, and particularly relates to the technical field of separating quality intelligent prediction, and through the construction of a load posture coupling space-time description set, the extraction of an axial unevenness coefficient and a load peak value offset, and the construction of a threshing intensity axial difference factor by combining a cylinder rotating speed, the threshing intensity axial difference factor is coupled with a pitch angle velocity to construct an inertial interference quantity, a vibration signal time-frequency decoupling method is used to separate low-frequency rigid body motion and high-frequency threshing characteristic components, a time series prediction model is used to perform distributed prediction on the entrainment loss rate of each axial section, a dynamic correction factor based on a measured deviation is constructed, a historical working condition fusion mechanism is used, online self-correction of the model is realized, a comprehensive quality index is compared with a preset threshold in a closed loop, and the cylinder rotating speed is adjusted in linkage, and adaptive control of threshing energy input is realized.

Description

Technical Field

[0001] This invention relates to the field of intelligent prediction technology for separation quality, and more specifically, to an intelligent prediction system for the threshing and separation quality of soybean combine harvesters. Background Technology

[0002] In the process of modern agricultural intelligence, intelligent prediction technology for the threshing and separating quality of soybean combine harvesters is considered a core means to achieve precise and loss-reducing harvesting. However, the complexity of the field operating environment poses a severe challenge to the reliability of this technology, especially in scenarios involving attitude disturbances caused by drastic terrain changes. When the harvester suddenly enters a deep ditch or crosses a field ridge from a stable driving position, the violent pitching or tilting of the machine body can instantly change the physical field distribution of the threshing system. This dynamic process brings profound physical challenges to existing intelligent prediction methods.

[0003] Threshing drums typically employ a long-shaft structure, and their separation performance highly depends on the uniformity of the axial distribution of the internal material. When the machine tilts, the crop grain mixture inside the drum rapidly accumulates towards the lower end under gravity, causing a momentary reconfiguration of the load distribution. This physical phenomenon directly undermines the lumped parameter assumption commonly used in existing prediction models, which assumes that the material state inside the drum is homogeneous and stable. At the moment of tilting, the higher end, due to material scarcity, faces the risk of excessive impact leading to a surge in breakage, while the lower end suffers from incomplete threshing and entrainment losses due to severe material accumulation. This renders traditional prediction models based on uniform field theory completely ineffective.

[0004] More complicated by the fact that the axial inertial force generated by the violent pitching motion further interferes with the dynamic characteristics of the separation process. This inertial force alters the trajectory of the grains as they sink through the straw layer and into the concave plate. Normal separation, which could have been accomplished by gravity and high-frequency vibration, may be forced back into the straw layer by the inertial impact and discharged through the straw outlet, causing unexpected hidden losses.

[0005] However, the vibration sensors configured in current mainstream prediction systems are mostly designed for high-frequency threshing characteristics, making it difficult to accurately capture and decouple such low-frequency, large-amplitude rigid body motion disturbances. Therefore, in scenarios with abrupt terrain changes, how to overcome the limitations of lumped parameter models and establish an intelligent sensing system capable of simultaneously identifying the coupling effect of attitude disturbances and separation mechanisms has become a key technical problem that urgently needs to be solved to improve the adaptive control accuracy of soybean combine harvesters. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an intelligent prediction system for the threshing and separating quality of soybean combine harvesters, in order to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] The intelligent prediction system for threshing and separating quality of soybean combine harvesters includes a data acquisition and description module, an axial difference calculation module, a disturbance decoupling analysis module, a loss distribution prediction module, a dynamic correction and evaluation module, and a closed-loop control output module.

[0009] The data acquisition and description module is used to acquire vehicle body attitude angle, attitude angular velocity, threshing drum speed, drum torque, and vibration acceleration signals collected from multiple vibration measuring points arranged along the drum axis. By sampling along the axial direction, the axial load distribution characteristics of the drum are constructed to form a load attitude coupled spatiotemporal description set.

[0010] The axial difference calculation module is used to extract the axial non-uniformity coefficient and load peak offset based on the load attitude coupled spatiotemporal description set, and calculate the axial difference factor of threshing intensity in combination with the drum speed.

[0011] The disturbance decoupling analysis module is used to construct the interference of inertial force on the grain settling trajectory based on the axial difference factor of threshing intensity and the pitch angular velocity of the vehicle body. Combined with the time-frequency analysis results of the vibration signal, the low-frequency rigid body motion component and the high-frequency threshing characteristic component are decoupled to obtain the inertial threshing coupling disturbance coefficient.

[0012] The loss distribution prediction module is used to predict the entrainment loss rate along each axial section of the drum based on the inertial threshing coupling disturbance coefficient and the real-time axial load distribution, and to obtain the predicted value of the entrainment loss rate distribution.

[0013] The dynamic correction and evaluation module is used to calculate the dynamic correction factor based on the deviation between the predicted value of the entrainment loss rate distribution and the actual value of the loss sensor at the drum outlet, and to obtain the overall threshing and separation quality comprehensive index by integrating historical working condition data.

[0014] The closed-loop control output module is used to compare the overall threshing and separation quality comprehensive index with the preset separation quality threshold. If the threshold is exceeded, the adjustment amount of the threshing drum speed is calculated.

[0015] In a preferred embodiment, the process of forming a load attitude coupled spatiotemporal description set is as follows:

[0016] Multi-source sensing units are deployed in the threshing drum area and body of the soybean combine harvester to continuously collect dynamic response parameters related to threshing and separation quality.

[0017] The dynamic response parameters include vehicle roll angle, pitch angle, attitude angular velocity, real-time rotational speed of the rollers, roller drive torque, and vibration acceleration signal.

[0018] All sensing units are connected to the vehicle data acquisition system and synchronized with the clock.

[0019] Set a uniform sampling frequency for the collected dynamic response parameters;

[0020] During harvester operation, within a fixed time window Sliding buffer is used for dynamic response parameters at each time step. The dynamic response parameters are preprocessed;

[0021] The preprocessed axial vibration sequence The corresponding drum speed and drum drive torque By fusing the data, a characteristic vector of the axial load distribution of the drum is constructed: , among which, the Load characterization value of measuring point This is a normalized combination of the effective value of vibration and torque. This represents the total number of axial positions.

[0022] The characteristic vector of the axial load distribution of the drum The load attitude coupled spatiotemporal description set is formed by mapping the roll angle, pitch angle and corresponding attitude angular velocity together.

[0023] In a preferred embodiment, the logic for obtaining the axial non-uniformity coefficient is as follows: extracting the axial load distribution feature vector of the drum from the load attitude coupled spatiotemporal description set. And calculate the standard deviation of the axial load distribution respectively. with the mean The axial non-uniformity coefficient is obtained by calculating the ratio of the standard deviation to the mean of the axial load distribution.

[0024] In a preferred embodiment, the logic for obtaining the load peak offset is as follows: from the roller axial load distribution feature vector Obtain load characterization values The maximum value, the load characterization value The load peak offset is obtained by calculating the difference between the maximum value and the average value.

[0025] In a preferred embodiment, the logic for obtaining the axial difference factor of threshing strength is as follows: based on the axial non-uniformity coefficient... and load peak offset The axial difference factor of threshing intensity was calculated by combining the drum rotation speed. The formula is as follows: ,in This is the axial non-uniformity coefficient. This represents the load peak offset. This refers to the drum rotation speed. This is the standard rotational speed.

[0026] In a preferred embodiment, the interference of inertial force on the grain settling trajectory is constructed based on the axial difference factor of threshing strength and the vehicle pitch angular velocity, and the calculation formula is as follows: ,in This represents the amount of inertial force that interferes with the grain settling trajectory. The axial difference factor of threshing strength. It represents the pitch angular velocity.

[0027] In a preferred embodiment, the process of obtaining the inertial de-granulation coupling perturbation coefficient is as follows:

[0028] Vibration acceleration signal Time-frequency analysis was performed, and short-time Fourier transform was used to obtain the time-frequency data for each moment. Spectral representation of the lower vibration acceleration signal ;

[0029] Based on a preset frequency range, the vibration acceleration signal is divided into a low-frequency rigid body motion component and a high-frequency threshing characteristic component. The specific separation rules are as follows: ,in Low-frequency components, For high-frequency components, and These represent the frequency range values ​​for low and high frequencies, respectively. The maximum frequency of the vibration acceleration signal;

[0030] After normalizing the interference of inertial force on the grain settling trajectory, including the low-frequency and high-frequency components, the inertial threshing coupling disturbance coefficient is calculated using the following formula: ,in The inertial de-granulation coupling perturbation coefficient is... , These are the preset weighting coefficients for the interference of inertial force on the grain settling trajectory and the ratio of low-frequency components to high-frequency components, respectively.

[0031] In a preferred embodiment, the process of obtaining the overall threshing and separation quality comprehensive index is as follows:

[0032] The difference between the predicted entrainment loss rate and the actual measured value of the loss sensor at the roller exit is calculated to obtain the loss rate deviation value.

[0033] The dynamic correction factor is calculated based on the loss rate deviation value, using the following formula: ,in For dynamic correction factors;

[0034] The overall threshing and separation quality comprehensive index is obtained by integrating historical operating data.

[0035] In a preferred embodiment, the overall threshing and separation quality comprehensive index is compared with a preset separation quality threshold. If the overall threshing and separation quality comprehensive index is greater than the separation quality threshold, it indicates that the separation quality exceeds the limit.

[0036] If the overall threshing and separation quality index is less than or equal to the separation quality threshold, then the separation quality has not exceeded the limit and the current state is maintained.

[0037] In a preferred embodiment, if the separation quality exceeds the limit, the threshing drum speed adjustment is calculated based on the difference between the overall threshing and separation quality comprehensive index and the separation quality threshold, using the following formula: ,in This refers to the adjustment amount of the threshing drum speed. The overall threshing and separation quality index. The preset separation quality threshold, This is the speed adjustment coefficient.

[0038] The technical effects and advantages of this invention are as follows:

[0039] 1. This invention addresses the problem that the accuracy of traditional threshing and separation quality prediction models decreases due to the imbalance in the axial distribution of material inside the drum and the increased inertial disturbance caused by sudden changes in the tilt and pitch attitude of the soybean combine harvester during operation in complex field terrain. By constructing a load-attitude coupled spatiotemporal description set that includes vehicle attitude parameters, drum operating parameters, and axial vibration characteristics, the axial load change of the material inside the drum is dynamically characterized. Furthermore, the axial difference factor of threshing intensity and the inertial threshing coupling disturbance coefficient are extracted to effectively identify the coupling effect between attitude disturbance and the threshing and separation process.

[0040] 2. This invention combines a time-series prediction model to predict the entrainment loss rate of each axial section of the drum, and uses the measured value of the outlet loss sensor for dynamic correction. Finally, it forms a comprehensive index of overall threshing and separation quality and drives the drum speed for closed-loop adjustment. This enables real-time prediction and adaptive control of threshing and separation quality under complex operating conditions, improves prediction accuracy and control response capability, reduces entrainment loss and grain breakage rate, and enhances the operational stability and loss reduction level of soybean combine harvesters. Attached Figure Description

[0041] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings;

[0042] Figure 1 This is a flowchart of the system according to an embodiment of the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] Example: The present invention provides, as follows Figure 1 The intelligent prediction system for threshing and separating quality of soybean combine harvester shown includes a data acquisition and description module, an axial difference calculation module, a disturbance decoupling analysis module, a loss distribution prediction module, a dynamic correction and evaluation module, and a closed-loop control output module.

[0045] The data acquisition and description module is used to acquire vehicle body attitude angle, attitude angular velocity, threshing drum speed, drum torque, and vibration acceleration signals collected from multiple vibration measuring points arranged along the drum axis. By sampling along the axial direction, the axial load distribution characteristics of the drum are constructed to form a load attitude coupled spatiotemporal description set.

[0046] The axial difference calculation module is used to extract the axial non-uniformity coefficient and load peak offset based on the load attitude coupled spatiotemporal description set, and calculate the axial difference factor of threshing intensity in combination with the drum speed.

[0047] The disturbance decoupling analysis module is used to construct the interference of inertial force on the grain settling trajectory based on the axial difference factor of threshing intensity and the pitch angular velocity of the vehicle body. Combined with the time-frequency analysis results of the vibration signal, the low-frequency rigid body motion component and the high-frequency threshing characteristic component are decoupled to obtain the inertial threshing coupling disturbance coefficient.

[0048] The loss distribution prediction module is used to predict the entrainment loss rate along each axial section of the drum based on the inertial threshing coupling disturbance coefficient and the real-time axial load distribution, and to obtain the predicted value of the entrainment loss rate distribution.

[0049] The dynamic correction and evaluation module is used to calculate the dynamic correction factor based on the deviation between the predicted value of the entrainment loss rate distribution and the actual value of the loss sensor at the drum outlet, and to obtain the overall threshing and separation quality comprehensive index by integrating historical working condition data.

[0050] The closed-loop control output module is used to compare the overall threshing and separation quality comprehensive index with the preset separation quality threshold. If the threshold is exceeded, the threshing drum speed adjustment amount is calculated and output to the actuator to achieve closed-loop control.

[0051] In this embodiment of the invention, the process of acquiring vehicle body attitude angle, attitude angular velocity, threshing drum rotation speed, drum torque, and vibration acceleration signals collected from multiple vibration measuring points arranged along the drum axis, and constructing the drum axial load distribution characteristics by sampling along the axial direction to form a load attitude coupled spatiotemporal description set is as follows:

[0052] Multi-source sensing units are deployed in the threshing drum area and key locations on the body of the soybean combine harvester to continuously collect dynamic response parameters related to threshing and separation quality during field operations and under changing terrain and attitude conditions; the sensing units include:

[0053] An inertial measurement unit installed near the vehicle's center of gravity is used to acquire the vehicle's roll angle in real time. Pitch angle and the corresponding attitude angular velocity , ;

[0054] The speed sensor and torque sensor mounted on the main shaft of the threshing drum are used to collect the real-time speed of the drum. and drum drive torque ;

[0055] Multiple piezoelectric vibration acceleration sensors, evenly spaced along the drum axis, are used to acquire data on the threshing drum shell and support at different axial positions. Vibration acceleration signal at the location ,in For the first Axial position, For the corresponding number Vibration acceleration signals at each axial position;

[0056] All sensing units are connected to the vehicle data acquisition system and clock-synchronized to ensure sampling consistency of each data source under a unified time reference.

[0057] A uniform sampling frequency is set for the collected dynamic response parameters so that the threshing process parameters form high-resolution continuous time-series data on the time axis.

[0058] Preferably, the sampling frequency is set to no less than 500Hz based on the drum speed and the main frequency range of the vibration signal, so as to fully preserve the high-frequency characteristics generated by the impact of material threshing and the low-frequency dynamics of vehicle body posture changes;

[0059] During harvester operation, within a fixed time window Sliding buffer is used for dynamic response parameters at each time step. The dynamic response parameters are preprocessed, including using a bandpass filter to filter out unrelated noise in the vibration signal and using a moving average method to smooth the instantaneous pulsations in the torque and flow signals.

[0060] The preprocessed axial vibration sequence The corresponding drum speed and drum drive torque By fusing the data, a characteristic vector of the axial load distribution of the drum is constructed: , among which, the Load characterization value of measuring point This is a normalized combination of the effective value of vibration and torque. This represents the total number of axial positions.

[0061] For example, load characterization value The calculation formula is as follows: ,in, For vibration acceleration signals within a time window The root mean square value of the discrete samples within the sample. , These are preset weighting coefficients for the root mean square value, the ratio of drum speed to drum drive torque, and so on.

[0062] It should be noted that, , The settings should be tailored to the specific circumstances. For example, an expert weighting method can be used, which involves inviting experts in relevant fields to determine the pre-defined weighting coefficients for each indicator through professional opinion surveys and comprehensive evaluations. , The initial value can be 0.5, 0.5;

[0063] The characteristic vector of the axial load distribution of the drum By jointly mapping the roll angle, pitch angle, and corresponding attitude angular velocity, a load-attitude coupled spatiotemporal description set is formed: ;

[0064] It should be noted that the load attitude coupling spatiotemporal description set not only records the load state of each section of the drum axis at each moment, but also couples the attitude disturbance information caused by terrain changes. Through time synchronization and spatial axial positioning, it establishes an explicit correlation between the dynamic response inside the threshing drum and the attitude change of the whole machine, and retains the spatiotemporal coupling characteristics of the load distribution along the axial direction evolving with attitude changes.

[0065] In this embodiment of the invention, the process of extracting the axial non-uniformity coefficient and load peak offset based on the load attitude coupled spatiotemporal description set, and calculating the axial difference factor of threshing intensity in combination with the drum rotation speed is as follows:

[0066] Extract the axial load distribution feature vector of the drum from the load attitude coupled spatiotemporal description set. And calculate the standard deviation of the axial load distribution respectively. with the mean The axial non-uniformity coefficient is obtained by calculating the ratio of the standard deviation to the mean of the axial load distribution.

[0067] For example, the axial non-uniformity coefficient The calculation formula is as follows: ,in, Characteristic vector of axial load distribution of the drum standard deviation Characteristic vector of axial load distribution of the drum The mean;

[0068] It should be noted that the axial non-uniformity coefficient It characterizes the uniformity of material distribution; the larger the value, the more uneven the material distribution.

[0069] From the characteristic vector of drum axial load distribution Obtain load characterization values The maximum value, the load characterization value The load peak offset is obtained by subtracting the maximum value from the mean value.

[0070] For example, load peak offset The calculation formula is as follows: ,in, Used to extract the characteristic vector of drum axial load distribution Obtain load characterization values The maximum value;

[0071] According to the axial non-uniformity coefficient and load peak offset The axial difference factor of threshing intensity was calculated by combining the drum rotation speed. The formula is as follows: ,in This is the axial non-uniformity coefficient. This represents the load peak offset. This refers to the drum rotation speed. The standard rotational speed indicates the reference rotational speed of the drum.

[0072] In this embodiment of the invention, based on the axial difference factor of threshing intensity and the vehicle pitch angular velocity, the interference of inertial force on the grain settling trajectory is constructed. Combining the time-frequency analysis results of the vibration signal, the low-frequency rigid body motion component and the high-frequency threshing characteristic component are decoupled to obtain the inertial threshing coupling disturbance coefficient.

[0073] In tilt and pitch states, the vehicle's movement generates inertial force interference, affecting the material's settling trajectory and threshing effect. Based on the axial difference factor of threshing intensity and the vehicle's pitch angular velocity, the interference of inertial force on the grain settling trajectory is constructed to represent the interference of inertial force caused by the vehicle's movement on the grain settling trajectory. The calculation formula is as follows: ,in This represents the amount of inertial force that interferes with the grain settling trajectory. The axial difference factor of threshing strength. It is the pitch angular velocity;

[0074] It should be noted that the amount of interference of inertial force on the grain settling trajectory It describes the motion displacement of materials due to inertia, reflecting the impact of vehicle body movement on materials;

[0075] In order to effectively distinguish between low-frequency rigid body motion and high-frequency threshing characteristics, the low-frequency and high-frequency components of the vibration signal are decoupled using the time-frequency analysis method of the vibration signal.

[0076] Vibration acceleration signal Time-frequency analysis was performed, and short-time Fourier transform was used to obtain the time-frequency data for each moment. Spectral representation of the lower vibration acceleration signal ;

[0077] Based on a preset frequency range, the vibration acceleration signal is divided into a low-frequency rigid body motion component and a high-frequency threshing characteristic component. The specific separation rules are as follows: ,in Low-frequency components, For high-frequency components, and These represent the frequency range values ​​for low and high frequencies, respectively. The maximum frequency of the vibration acceleration signal;

[0078] It should be noted that the low-frequency components mainly reflect the rigid body motion of the vehicle body, which is mainly caused by the pitch and roll of the vehicle body; while the high-frequency components reflect the threshing characteristics caused by the impact of materials and the rotation of the drum during the threshing process.

[0079] After normalizing the interference of inertial force on the grain settling trajectory, including the low-frequency and high-frequency components, the inertial threshing coupling disturbance coefficient is calculated using the following formula: ,in The inertial de-granulation coupling perturbation coefficient is... , These are the preset weighting coefficients for the interference of inertial force on the grain settling trajectory and the ratio of low-frequency components to high-frequency components, respectively.

[0080] It should be noted that, , The settings should be tailored to the specific circumstances. For example, an expert weighting method can be used, which involves inviting experts in relevant fields to determine the pre-defined weighting coefficients for each indicator through professional opinion surveys and comprehensive evaluations. , The initial value can be 0.5, 0.5;

[0081] It should also be noted that the inertial threshing coupling disturbance coefficient comprehensively reflects the direct interference of pitch inertial force on grain settling and the coupling effect of the rigid body vibration of the whole machine caused by attitude change on the measurement of the threshing process.

[0082] In this embodiment of the invention, the process of predicting the entrainment loss rate along each axial section of the drum using a time-series prediction model based on the inertial threshing coupling disturbance coefficient and the real-time axial load distribution, and obtaining the predicted value of the entrainment loss rate distribution, is as follows:

[0083] The inertial de-granulation coupling perturbation coefficient With the characteristic vector of the axial load distribution of the drum The input feature matrix is ​​combined into a multidimensional input feature matrix, which is then processed by a pre-trained time-series prediction model. The time-series prediction model internally uses a gating mechanism to capture the temporal dependency between load and disturbance, and outputs predicted entrainment loss rates for each axial segment of the roller at future times. ,in This is the output function of the time series prediction model. For the first Axial position, For the first Predicted entrainment loss rate at each axial position;

[0084] It should be noted that the time series prediction model can use existing recurrent neural network (RNN) or long short-term memory network (LSTM) models. The entrainment loss rate refers to the proportion of threshed grains that failed to pass through the concave sieve and were entrained by straw or debris and discharged through the straw discharge port to the total yield.

[0085] For example, the structure of a time series prediction model is as follows: ,in The mapping function representing the time series prediction model. This is the predicted value for the entrainment loss rate.

[0086] In this embodiment of the invention, the process of calculating a dynamic correction factor based on the deviation between the predicted entrainment loss rate distribution and the actual loss rate measured by the loss sensor at the drum outlet, and then integrating historical operating data to obtain the overall threshing and separation quality comprehensive index, is as follows:

[0087] The difference between the predicted entrainment loss rate and the actual measured value from the loss sensor at the roller exit is calculated to obtain the loss rate deviation value. The calculation formula is as follows: ,in This is the deviation value of the loss rate. For the first Predicted entrainment loss rate at each axial position. This is the measured value at the roller exit point;

[0088] The dynamic correction factor is calculated based on the loss rate deviation value, using the following formula: ,in For dynamic correction factors;

[0089] And integrate historical operating data The overall threshing and separation quality comprehensive index was obtained: ,in The overall threshing and separation quality index. For a moment The obtained dynamic correction factor This is the average of historical operating data. , They are time points The dynamic correction factor and the preset weighting coefficient of the historical operating condition data mean;

[0090] It should be noted that, , The settings should be tailored to the specific circumstances. For example, an expert weighting method can be used, which involves inviting experts in relevant fields to determine the pre-defined weighting coefficients for each indicator through professional opinion surveys and comprehensive evaluations. , The initial value can be 0.5 or 0.5.

[0091] In this embodiment of the invention, the process of comparing the overall threshing and separation quality comprehensive index with a preset separation quality threshold, and calculating the adjustment amount of the threshing drum speed if the threshold is exceeded, and outputting it to the actuator to achieve closed-loop control is as follows:

[0092] The overall threshing and separation quality comprehensive index is compared with a preset separation quality threshold. If the overall threshing and separation quality comprehensive index is greater than the separation quality threshold, it indicates that the separation quality exceeds the limit.

[0093] If the overall threshing and separation quality index is less than or equal to the separation quality threshold, then the separation quality has not exceeded the limit and the current state is maintained.

[0094] If the speed exceeds the limit, the rotation speed of the threshing drum needs to be adjusted. To control energy transfer and material separation efficiency during the threshing process, the formula for calculating the adjustment amount of the threshing drum speed is as follows, based on the difference between the overall threshing and separation quality comprehensive index and the separation quality threshold: ,in This refers to the adjustment amount of the threshing drum speed. The overall threshing and separation quality index. The preset separation quality threshold, This is the speed adjustment coefficient, representing the sensitivity of speed adjustment when the mass exceeds the limit;

[0095] It should be noted that the above The rotational speed adjustment ratio coefficient, calibrated through bench tests and field comparison tests, is used to characterize the sensitivity of drum rotational speed adjustment when the overall threshing and separation quality comprehensive index exceeds the limit.

[0096] To ensure a balance between stability and response speed during system adjustment, the following... The value of must meet the following range requirement: 0.1≤ ≤5.00;

[0097] Preferably, under the operating conditions of a conventional feed rate of 3–6 kg / s and a drum rated speed of 600–900 r / min, the... The value range is: 0.5≤ ≤1.50;

[0098] when When the value is less than 0.1, the speed adjustment response is too slow, which may lead to a persistent state of excessive mass, affecting the separation effect; when... When the value is greater than 5.0, it is easy to cause overshoot in the drum speed adjustment, resulting in system oscillation or a sudden increase in the breakage rate, which is not conducive to stable control.

[0099] This invention addresses the problem of decreased accuracy in traditional threshing and separation quality prediction models caused by abrupt changes in the lateral and pitch attitudes of soybean combine harvesters during operation in complex field terrain. This imbalance in axial material distribution within the drum and increased inertial disturbances lead to a decline in accuracy. The invention constructs a load-attitude coupled spatiotemporal description set, incorporating vehicle attitude parameters, drum operating parameters, and axial vibration characteristics. This dynamically characterizes the axial load changes of the material inside the drum and extracts the threshing intensity axial difference factor and the inertial threshing coupling disturbance coefficient, effectively identifying the coupling effect between attitude disturbances and the threshing and separation process. Combined with a time-series prediction model, the invention predicts the entrainment loss rate in each axial section of the drum and uses measured values ​​from the outlet loss sensor for dynamic correction. Finally, it generates a comprehensive threshing and separation quality index and drives closed-loop adjustment of the drum speed. This enables real-time prediction and adaptive control of threshing and separation quality under complex operating conditions, improving prediction accuracy and control response, reducing entrainment losses and grain breakage rates, and enhancing the operational stability and loss reduction of soybean combine harvesters.

[0100] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0101] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0102] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0103] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0104] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An intelligent prediction system for the threshing and separating quality of soybeans in a combine harvester, characterized in that: It includes a data acquisition and description module, an axial difference calculation module, a disturbance decoupling analysis module, a loss distribution prediction module, a dynamic correction and evaluation module, and a closed-loop control output module; The data acquisition and description module is used to acquire vehicle body attitude angle, attitude angular velocity, threshing drum speed, drum torque, and vibration acceleration signals collected from multiple vibration measuring points arranged along the drum axis. By sampling along the axial direction, the axial load distribution characteristics of the drum are constructed to form a load attitude coupled spatiotemporal description set. The axial difference calculation module is used to extract the axial non-uniformity coefficient and load peak offset based on the load attitude coupled spatiotemporal description set, and calculate the axial difference factor of threshing intensity in combination with the drum speed. The disturbance decoupling analysis module is used to construct the interference of inertial force on the grain settling trajectory based on the axial difference factor of threshing intensity and the pitch angular velocity of the vehicle body. Combined with the time-frequency analysis results of the vibration signal, the low-frequency rigid body motion component and the high-frequency threshing characteristic component are decoupled to obtain the inertial threshing coupling disturbance coefficient. The loss distribution prediction module is used to predict the entrainment loss rate along each axial section of the drum based on the inertial threshing coupling disturbance coefficient and the real-time axial load distribution, and to obtain the predicted value of the entrainment loss rate distribution. The dynamic correction and evaluation module is used to calculate the dynamic correction factor based on the deviation between the predicted value of the entrainment loss rate distribution and the actual value of the loss sensor at the drum outlet, and to obtain the overall threshing and separation quality comprehensive index by integrating historical working condition data. The closed-loop control output module is used to compare the overall threshing and separation quality comprehensive index with the preset separation quality threshold. If the threshold is exceeded, the adjustment amount of the threshing drum speed is calculated. The process of forming a load attitude coupled spatiotemporal description set is as follows: Multi-source sensing units are deployed in the threshing drum area and body of the soybean combine harvester to continuously collect dynamic response parameters related to threshing and separation quality. The dynamic response parameters include vehicle roll angle, pitch angle, attitude angular velocity, real-time rotational speed of the rollers, roller drive torque, and vibration acceleration signal. All sensing units are connected to the vehicle data acquisition system and synchronized with the clock. Set a uniform sampling frequency for the collected dynamic response parameters; During harvester operation, within a fixed time window Sliding buffer is used for dynamic response parameters at each time step. The dynamic response parameters are preprocessed; The preprocessed axial vibration sequence The corresponding drum speed and drum drive torque By fusing the data, a characteristic vector of the axial load distribution of the drum is constructed: , among which, the Load characterization value of measuring point This is a normalized combination of the effective value of vibration and torque. This represents the total number of axial positions. The characteristic vector of the axial load distribution of the drum The load attitude coupled spatiotemporal description set is formed by mapping the roll angle, pitch angle and corresponding attitude angular velocity together.

2. The intelligent prediction system for threshing and separating quality of soybean combine harvesters according to claim 1, characterized in that: The logic for obtaining the axial non-uniformity coefficient is as follows: extract the axial load distribution feature vector of the drum from the load attitude coupled spatiotemporal description set. And calculate the standard deviation of the axial load distribution respectively. with the mean The axial non-uniformity coefficient is obtained by calculating the ratio of the standard deviation to the mean of the axial load distribution.

3. The intelligent prediction system for threshing and separating quality of soybean combine harvesters according to claim 1, characterized in that: The logic for obtaining the load peak offset is as follows: from the roller axial load distribution feature vector Obtain load characterization values The maximum value, the load characterization value The load peak offset is obtained by calculating the difference between the maximum value and the average value.

4. The intelligent prediction system for threshing and separating quality of soybean combine harvesters according to claim 3, characterized in that: The logic for obtaining the axial difference factor of threshing strength is as follows: based on the axial non-uniformity coefficient. and load peak offset The axial difference factor of threshing intensity was calculated by combining the drum rotation speed. The formula is as follows: ,in This is the axial non-uniformity coefficient. This represents the load peak offset. This refers to the drum rotation speed. This is the standard rotational speed.

5. The intelligent prediction system for threshing and separating quality of soybean combine harvesters according to claim 4, characterized in that: Based on the axial difference factor of threshing strength and the vehicle pitch velocity, the interference of inertial force on the grain settling trajectory is constructed, and the calculation formula is as follows: ,in This represents the amount of inertial force that interferes with the grain settling trajectory. The axial difference factor of threshing strength. It represents the pitch angular velocity.

6. The intelligent prediction system for threshing and separating quality of soybean combine harvesters according to claim 5, characterized in that: The process of obtaining the inertial de-granulation coupling perturbation coefficient is as follows: Vibration acceleration signal Time-frequency analysis was performed, and short-time Fourier transform was used to obtain the time-frequency data for each moment. Spectral representation of the lower vibration acceleration signal ; Based on a preset frequency range, the vibration acceleration signal is divided into a low-frequency rigid body motion component and a high-frequency threshing characteristic component. The specific separation rules are as follows: ,in Low-frequency components, For high-frequency components, and These represent the frequency range values ​​for low and high frequencies, respectively. The maximum frequency of the vibration acceleration signal; After normalizing the interference of inertial force on the grain settling trajectory, including the low-frequency and high-frequency components, the inertial threshing coupling disturbance coefficient is calculated using the following formula: ,in The inertial de-granulation coupling perturbation coefficient is... , These are the preset weighting coefficients for the interference of inertial force on the grain settling trajectory and the ratio of low-frequency components to high-frequency components, respectively.

7. The intelligent prediction system for threshing and separating quality of soybean combine harvesters according to claim 1, characterized in that: The process of obtaining the overall threshing and separation quality comprehensive index is as follows: The difference between the predicted entrainment loss rate and the actual measured value of the loss sensor at the roller exit is calculated to obtain the loss rate deviation value. The dynamic correction factor is calculated based on the loss rate deviation value, using the following formula: ,in For dynamic correction factors; The overall threshing and separation quality comprehensive index is obtained by integrating historical operating data.

8. The intelligent prediction system for threshing and separating quality of soybean combine harvesters according to claim 7, characterized in that: The overall threshing and separation quality comprehensive index is compared with a preset separation quality threshold. If the overall threshing and separation quality comprehensive index is greater than the separation quality threshold, it indicates that the separation quality exceeds the limit. If the overall threshing and separation quality index is less than or equal to the separation quality threshold, then the separation quality has not exceeded the limit and the current state is maintained.

9. The intelligent prediction system for threshing and separating quality of soybean combine harvesters according to claim 8, characterized in that: If the separation quality exceeds the limit, the threshing drum speed adjustment is calculated based on the difference between the overall threshing and separation quality comprehensive index and the separation quality threshold. The formula is as follows: ,in This refers to the adjustment amount of the threshing drum speed. The overall threshing and separation quality index. The preset separation quality threshold, This is the speed adjustment coefficient.

Images