A method and apparatus for feed monitoring with light loss detection

Abstract

The application provides a kind of light loss detection feeding amount monitoring method and device, it is related to feeding amount monitoring technical field, first in the stem state calibration reference light intensity value;In operation, the light intensity signal when stem sheltering light transmission seam is collected in real time, and the chain harrow transmission speed is obtained synchronously. Again, the light intensity signal is calculated by using the integral model, and the average light loss rate in the unit length is calculated. This step significantly improves the stability and accuracy of measurement. Finally, based on the pre-calibrated light loss rate and the fitting curve of feeding amount, the real-time feeding amount is calculated and automatic control is realized. The application overcomes the defects of traditional mechanical sensor being easily disturbed by vibration and visual scheme relying on light, has the advantages of strong anti-interference ability and all-weather working, and provides reliable technical support for precise control of combine harvester feeding amount and efficient and low-loss operation.

Description

Technical Field

[0001] This invention relates to the field of feed volume monitoring technology, specifically to a feed volume monitoring method and apparatus for detecting light loss. Background Technology

[0002] Feed rate is one of the key parameters in the operation of a grain combine harvester, closely related to the performance of various components and directly affecting harvesting efficiency and grain loss rate. In actual operation, if the feed rate exceeds the rated processing capacity of the threshing and separating system, it can easily lead to malfunctions such as chute blockage, drum entanglement, and even damage to critical components. Conversely, if the feed rate is consistently below the machine's optimal operating range, the harvester's performance will not be fully utilized, resulting in wasted power and fuel resources. Therefore, achieving real-time and accurate monitoring of the feed rate is an important prerequisite for ensuring the efficient, low-loss, and reliable operation of the combine harvester.

[0003] Currently, the monitoring technology for combine harvester feed rate mainly falls into two categories. One category is the traditional approach based on mechanical sensors, such as installing torque sensors at the conveyor bridge. This involves detecting torque changes during stalk transport and combining this with the machine's forward speed to indirectly estimate the feed rate. However, combine harvesters generate continuous and severe vibrations under harsh field conditions. Mechanical sensors, such as torque sensors, are highly susceptible to vibration interference in this environment, resulting in low signal-to-noise ratios, large measurement errors, mechanical wear, and data lag, severely impacting the accuracy and stability of monitoring. To overcome the shortcomings of mechanical sensors, emerging technologies are adopting non-contact detection methods based on machine vision. For example, depth cameras or vision cameras are used to capture images of the crop in front of the harvester, analyzing plant characteristics and using deep learning models to predict the feed rate. While this method avoids mechanical contact, its performance is highly dependent on ambient lighting conditions; the recognition effect drops sharply in low light, high light, or shadow conditions. Meanwhile, image processing algorithms are complex and require high computing power. They are also difficult to overcome the impact of field operation vibrations on image clarity, resulting in high system costs, insufficient reliability, and lack of ability to work stably around the clock.

[0004] In summary, existing feed rate monitoring technologies, whether mechanical sensors susceptible to vibration interference or vision-based solutions limited by lighting and computing power, struggle to achieve cost-effective, stable, reliable, and all-weather accurate measurement in the complex and variable operating environments of combine harvesters. This has become a technological bottleneck restricting further improvements in the intelligence level and operational quality of combine harvesters. Therefore, there is an urgent need for a novel monitoring principle and technological approach that can directly, in-situ, and non-contactly sense crop throughput during the conveying process, effectively resisting field interference such as vibration and lighting, thereby providing a reliable basis for precise feed rate control and intelligent overall machine regulation.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide an optimized configuration method and apparatus for a magnetic reactive dynamic voltage restorer to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A method for monitoring feed rate to detect light loss, comprising the following steps: Step 1: With no stalks passing through the conveying trough of the combine harvester, turn on the light source set at the bottom of the conveying trough. The light shines through the light-transmitting slit at the bottom of the conveying trough onto the photometric sensor group directly above. Measure and store the voltage value output by the photometric sensor group at this time as the reference light intensity value. Step 2: During the harvesting of the target crop by the combine harvester, the real-time light intensity signal output by the photometric sensor group when the crop stalk passes through the light-transmitting slit is collected in real time, and the real-time transmission speed of the chain rake is obtained by the chain rake proximity sensor, so as to calculate the chain rake real-time speed ratio by combining it with the preset standard chain rake transmission speed. Step 3: Based on the reference light intensity value and the real-time light intensity signal, calculate the average light loss rate of the crop stem per unit length. Step 4: Based on the fitting curve of light loss rate of target crop and feed rate at standard chain harrow transmission speed, determine the standard feed rate at average light loss rate, and calculate the real-time feed rate by combining the real-time speed ratio of chain harrow. Step 5: Compare the real-time feed rate with the preset rated feed rate threshold, and generate a control signal based on the comparison result to adjust the travel speed or header height of the combine harvester, thereby achieving real-time control of the feed rate.

[0008] Furthermore, the light source, the light-transmitting slit, and the photometric sensor group constitute a closed optical sensing unit; the light-transmitting slit is a through-slot with a width of 1.5 cm opened at the bottom of the conveying trough, used to constrain the light emitted by the light source into a narrow light band; the photometric sensor group consists of multiple photometric sensors arranged in a uniform distribution directly above the light-transmitting slit, used to receive the light passing through the light-transmitting slit; the photometric sensor includes a cosine correction plate, a silicon photodiode, and a signal conditioning circuit arranged sequentially along the optical path; the cosine correction plate is used to correct the incident light angle according to Lambert's cosine law; the signal conditioning circuit is a transimpedance amplifier circuit, used to convert the photocurrent signal generated by the silicon photodiode into a voltage signal output.

[0009] Furthermore, acquiring the real-time light intensity signal, real-time transmission speed, and the ratio of the chain rake's real-time speed specifically includes: The photometric sensor array continuously acquires light intensity signals at a sampling frequency of not less than 1kHz and outputs corresponding voltage values; simultaneously, the chain rake proximity sensor monitors the passing status of the chain rake in real time and outputs corresponding level signals. When the chain rake proximity sensor outputs a low-level signal indicating that the chain rake is approaching, it is determined that the light-transmitting slit is completely blocked by the chain rake at this time. The light intensity signal collected by the photometric sensor is an invalid signal, and the light intensity signal in this period is marked as invalid and not used. The moment when the output level signal of the chain rake proximity sensor changes from low to high is taken as the start of the effective detection interval, and the moment when the output level signal changes from high to low is taken as the end of the effective detection interval. The real-time light intensity signal is taken from the photometric sensor data within the effective detection interval. The chain rake proximity sensor monitors the time interval between the transition from a low-level signal to a high-level signal between two adjacent rake teeth, and calculates the ratio of the fixed distance between the two adjacent rake teeth to the time interval. This ratio is the real-time transmission speed of the chain rake. The real-time speed ratio of the chain rake is the ratio of the real-time transmission speed to the preset standard chain rake transmission speed.

[0010] Further, calculating the average optical loss rate specifically includes: The average optical loss rate within the detection interval is calculated using an integral model, and its mathematical model function is as follows: in, Indicates the average optical loss rate. This represents the real-time light intensity signal, where a and b represent the start and end points of the effective detection interval, respectively, and t represents the time variable. The effective detection range is defined as one effective detection cycle of the chain rake. Specifically, the starting point corresponds to the instant when a chain rake completely leaves the light-transmitting slit and the photometric sensor begins to detect the crop stem; the ending point corresponds to the instant when the next chain rake reaches and begins to block the light-transmitting slit and the photometric sensor is about to be shielded; the effective detection range is the length of one chain rake operation cycle.

[0011] Furthermore, the fitting curve between the light loss rate and the feed amount is established in the following manner: During the experimental calibration phase, the combine harvester was controlled to run at a constant speed at the standard chain harrow transmission speed, and data points were obtained under multiple different feeding conditions by changing the feed amount of the target crop. For each data point, record the actual feed amount obtained by weighing and calculate the corresponding average optical loss rate. The obtained data points of actual feed rate and average light loss rate are organized to form a data point set. The least squares method is used for curve fitting to establish a quantitative functional relationship between the light loss rate and the feed rate. The quantitative functional relationship is a quadratic polynomial model, and its expression is: in, This is the standard feed rate at the standard chain rake conveyor speed. These are the crop-specific fitting coefficients for different target crops, obtained through curve fitting. Based on the real-time conveying speed of the chain rake, the specific formula for calculating the real-time feed rate is as follows: Where Q represents the real-time feed volume, This represents the real-time speed ratio of the chain rake.

[0012] Furthermore, the aforementioned The three crop-specific fitting coefficients are calibrated and stored separately for soybeans, rice, or wheat. When the combine harvester switches to harvest a different crop, the corresponding fitting coefficients are called to update the quantitative function relationship.

[0013] Furthermore, a rated feed rate threshold is set, and the difference between the real-time feed rate and the rated feed rate threshold is calculated and used as the deviation. If the deviation is greater than the preset first positive threshold, it is determined to be overfeeding, and a control signal is generated to reduce the real-time feeding amount. If the deviation is less than the preset first negative threshold, it is determined that the feed is insufficient, and a control signal is generated to increase the real-time feed amount. If the deviation is within the preset range of the first negative threshold and the first positive threshold, it is determined that the feed amount is appropriate, and a control signal is generated to maintain the current working state. The control signal is configured to directly control the actuator of the combine harvester, specifically as follows: the control signal for reducing the real-time feed rate is configured to perform at least one of the following operations, namely, reducing the travel speed of the combine harvester or increasing the header height; the control signal for increasing the real-time feed rate is configured to perform at least one of the following operations, namely, increasing the travel speed of the combine harvester or reducing the header height.

[0014] The present invention also provides a feed rate monitoring device for light loss detection, wherein the feed rate monitoring device for light loss detection is used to implement the above-mentioned feed rate monitoring method for light loss detection, comprising: The system calibration and self-calibration module is used to turn on the light source set at the bottom of the conveying trough when there are no stalks passing through the conveying trough of the combine harvester. The light shines through the light-transmitting slit at the bottom of the conveying trough onto the photometric sensor group directly above, and measures and stores the voltage value output by the photometric sensor group at this time as the reference light intensity value. The signal acquisition and preprocessing module is used to acquire the real-time light intensity signal output by the photometric sensor group when the crop stalk passes through the light-transmitting slit during the harvesting operation of the combine harvester. It also acquires the real-time transmission speed of the chain rake through the chain rake proximity sensor and calculates the real-time speed ratio of the chain rake by combining it with the preset standard chain rake transmission speed. The light loss rate analysis module is used to calculate the average light loss rate of crop stems per unit length based on the reference light intensity value and the real-time light intensity signal. The feed rate mapping and correction module is used to determine the standard feed rate under the average light loss rate based on the fitting curve of the light loss rate of the target crop and the feed rate under the standard chain harrow transmission speed, and to calculate the real-time feed rate by combining the real-time speed ratio of the chain harrow. The decision and control module is used to compare the real-time feed rate with the preset rated feed rate threshold, and generate control signals based on the comparison results to adjust the travel speed or header height of the combine harvester, thereby achieving real-time control of the feed rate.

[0015] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention successfully overcomes the inherent defects of traditional mechanical sensors and vision solutions in monitoring the feed rate of combine harvesters. By adopting a non-contact light loss detection principle of light source-light transmission slit-photometer sensor group, the degree of stalk shading is directly measured in situ within the conveying trough. This invention fundamentally avoids signal interference and wear problems caused by severe field vibrations to mechanical torque sensors. At the same time, the closed optical path design effectively isolates the influence of external ambient light changes, ensuring high stability and reliability of monitoring data in all-weather operating scenarios.

[0016] Furthermore, based on the strong linear correlation between light loss rate and stalk density, combined with a simple mathematical model, this invention can quickly and accurately calculate the real-time feed rate. This not only reduces reliance on high-performance edge computing devices and significantly lowers system costs, but also enables real-time and precise monitoring of the feed rate. It provides an immediate and reliable basis for combine harvesters to automatically adjust their forward speed or header height according to the workload, effectively preventing material congestion or resource waste and ensuring efficient and low-loss operation of the machine within its optimal performance range. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 This is a schematic diagram of the fitting curve between the average light loss rate and the standard feed amount of the present invention; Figure 3 This is a schematic diagram of the conventional residuals for the average optical loss rate of this invention; Figure 4 This is a schematic diagram of the device structure of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0019] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0020] Example: Please see Figures 1 to 3 The present invention provides a technical solution: A method for monitoring feed rate to detect light loss, comprising the following steps: Step 1: With no stalks passing through the conveyor trough of the combine harvester, turn on the light source set at the bottom of the conveyor trough. The light shines through the light-transmitting slit at the bottom of the conveyor trough onto the photometric sensor group directly above. Measure and store the voltage value output by the photometric sensor group at this time as the reference light intensity value.

[0021] In this embodiment, the light source, the light-transmitting slit, and the photometric sensor group constitute a closed optical sensing unit. The light source, the light-transmitting slit, the photometric sensor group, and the chain rake proximity sensor are deployed on the same cross-sectional plane. The photometric sensor group has four channels evenly distributed above the light-transmitting slit. The chain rake proximity sensor is located on the side, flush with the end face of the chain rake when passing through. Preferably, the light source is a monochromatic LED light source, such as an LED emitting a specific wavelength (e.g., 850nm infrared light). This can effectively suppress interference from ambient visible light and reduce the impact of crop surface color differences on the measurement results.

[0022] The light-transmitting slit is located at the bottom of the conveying trough, directly above the light source. This slit is a 1.5 cm wide through-slot. The core function of this design is to constrain the scattered light emitted by the light source into a narrow, regularly shaped light band. This light band spans the width of the conveying trough, thus ensuring that crop stems can be effectively detected regardless of their location within the trough as they pass through.

[0023] The photometric sensor group consists of multiple photometric sensors arranged in a uniform distribution directly above the light-transmitting slit to ensure complete reception of light passing through the slit. Each individual photometric sensor is arranged sequentially along the optical path: Cosine corrector: Its function is to correct the angle of incident light based on Lambert's cosine law. Because the shape of the stem is irregular when it passes through, light may enter the sensor at different angles. Without correction, the measurement value will be too low. The cosine corrector, through its special optical design, ensures that regardless of the angle of light incidence, the sensor's response value is primarily proportional to the light intensity and inversely proportional to the cosine of the incident angle, thus significantly improving measurement accuracy under complex passing postures.

[0024] Silicon photodiode: As the core photoelectric conversion element, it converts the received light signal into a weak current signal (photocurrent).

[0025] Signal conditioning circuit: specifically, a transimpedance amplifier (TIA). This circuit converts the weak photocurrent signal generated by the silicon photodiode into a voltage signal that is easily processed by the subsequent system. The choice of the transimpedance amplifier is crucial because it provides low input impedance, can quickly respond to changes in photocurrent, and has good linearity and signal-to-noise ratio, making it suitable for high-speed, changing signal acquisition environments.

[0026] Step 2: During the harvesting of the target crop by the combine harvester, the real-time light intensity signal output by the photometric sensor group when the crop stalk passes through the light-transmitting slit is collected in real time, and the real-time transmission speed of the chain rake is obtained by the chain rake proximity sensor. The chain rake real-time speed ratio is calculated by combining it with the preset standard chain rake transmission speed.

[0027] This step is the core data acquisition environment for real-time monitoring of feed intake. Its purpose is to simultaneously acquire two key physical quantities: one is the light intensity signal reflecting the shading of the stalks, and the other is the transmission speed of the chain rake that determines the data normalization process.

[0028] In this embodiment, obtaining the ratio of the real-time light intensity signal, real-time transmission speed, and chain rake real-time speed specifically includes: The photometric sensor array continuously acquires light intensity signals and outputs corresponding voltage values ​​at a sampling frequency of no less than 1 kHz. Such a high sampling frequency is used to ensure that instantaneous light intensity fluctuations caused by crop stalks that pass rapidly between the rake teeth and may be unevenly distributed are captured. This avoids losing crucial occlusion information due to a low sampling rate, laying the foundation for subsequent calculations of the high-fidelity average light loss rate.

[0029] Simultaneously, the chain rake proximity sensor monitors the passing status of the chain rake in real time and outputs a corresponding level signal. In a preferred embodiment, the chain rake proximity sensor can be an inductive proximity switch, a Hall effect sensor, or a photoelectric sensor. When it detects the metal rake teeth approaching, it outputs a low-level signal; when the rake teeth move away and the detection area is unobstructed, it outputs a high-level signal. This level signal is a key reference for distinguishing valid and invalid data.

[0030] Before being used for calculation, the real-time light intensity signal needs to undergo validity verification and preprocessing, specifically including the following sub-steps: When the chain rake proximity sensor outputs a low-level signal indicating the chain rake's approach, it is determined that the light-transmitting slit is completely blocked by the chain rake. The light intensity signal collected by the photometer at this time contains only extremely low noise, unrelated to the passage of the crop stalk. Therefore, the system's internal microprocessor (such as an MCU or DSP) generates an invalid data flag, binding it to all light intensity voltage values ​​collected during that period. In subsequent data processing, all data points carrying this flag will be skipped and not used. This mechanism fundamentally eliminates the periodic interference of the chain rake itself on optical measurements.

[0031] The effective detection interval begins when the output signal of the chain harrow proximity sensor changes from low to high (rising edge), corresponding to the instant the end of a harrow tooth leaves the light-transmitting slit. The effective detection interval ends when the output signal changes from high to low (falling edge), corresponding to the instant the front of the next harrow tooth begins to enter and block the light-transmitting slit. The real-time light intensity signal is strictly taken from the valid photometric sensor data within the time window defined by these two moments. This window precisely corresponds to the entire process of a chain harrow tooth (the harrow window) completely passing through the light-transmitting slit, ensuring that the collected light intensity changes are purely caused by the crop stalks carried within that harrow window.

[0032] Due to the severe vibrations and complex electromagnetic interference in the operating environment of combine harvesters, the raw voltage signal directly acquired from the photometric sensor array may contain high-frequency noise and random interference pulses. Therefore, digital filtering preprocessing is required before using the light intensity signal within the effective detection range for calculation. As a preferred implementation, a moving average filter or a low-pass digital filter (e.g., a Butterworth low-pass filter with a cutoff frequency of 100Hz) can be used to smooth the light intensity signal within the effective range. For example, a moving average filter can calculate the arithmetic mean of the voltage values ​​at the current point and N points before and after it (e.g., N=5) to replace the original value at that point, thereby effectively suppressing random noise and smoothing the signal curve without significantly altering the macroscopic light intensity variation trend caused by stalk shading.

[0033] The time interval between the transition of two adjacent rake teeth from a low-level signal to a high-level signal (i.e., two consecutive rising edges) is monitored by a proximity sensor. Since the spacing between adjacent rake teeth on the chain rake is a fixed mechanical design parameter (e.g., 100 mm), the real-time transmission speed of the chain rake can be calculated using the following formula: Where V represents the real-time transmission speed, D represents the spacing, and T represents the time interval length. This calculation method is direct and reliable, and can respond in real time to fluctuations in the chain rake speed caused by changes in the forward speed of the combine harvester or the hydraulic system driving the chain rake.

[0034] The real-time speed ratio of the chain rake is determined by the following formula: in, This represents the real-time speed ratio of the chain rake. The standard chain rake conveying speed is preset.

[0035] The standard chain rake transmission speed This is not a theoretically derived value, but an experimentally calibrated value based on the optimal operational effect on the target crop. Its specific value should be determined in the following way: When designing and finalizing a combine harvester or optimizing operating parameters for a specific crop (such as soybeans), field trials are conducted by performing harvesting operations at different chain harrow speeds under rated operating conditions and a set feed rate. Simultaneously, operational quality indicators such as grain breakage rate, unthreshed loss, entrainment loss, and power consumption are evaluated. Ultimately, the chain harrow speed that achieves the optimal overall operational quality indicators (e.g., lowest total loss rate and reasonable power consumption) is selected and set as the standard chain harrow speed for this type of harvester for this crop. The rationale for this value is that it anchors the subsequent feed rate calculation to the optimal operating point designed for the machine itself. The speed ratio P is essentially a normalization factor; when the real-time transmission speed equals the standard speed... When P=1, the calculated feed rate does not require speed correction; when V deviates... When P≠1, it can linearly compensate for the feed rate measurement results, thereby eliminating the systematic error caused by changes in operating speed to the optical measurement method based on the fixed rake window, and ensuring the consistency and accuracy of the feed rate monitoring results under different operating speeds.

[0036] Step 3: Based on the reference light intensity value and the real-time light intensity signal, calculate the average light loss rate of the crop stem per unit length.

[0037] In this embodiment, calculating the average optical loss rate specifically includes: Necessary preprocessing is performed before applying the real-time light intensity signal to further improve signal quality. In a preferred embodiment of this invention, the preprocessing operations include, but are not limited to: Digital filtering: For example, a low-pass filter (such as a Butterworth low-pass filter or a moving average filter) can be used to smooth the signal. Since the passage of crop stalks is a relatively low-frequency physical process, while noise typically exhibits high-frequency characteristics, low-pass filtering can effectively suppress high-frequency noise, smooth the signal curve, and avoid losing useful information. The cutoff frequency of the filter can be set according to the maximum operating speed of the chain harrow and the typical passage characteristics of the stalks; for example, it can typically be set in the range of 100Hz to 500Hz.

[0038] Outlier Removal: In the data stream, instantaneous spikes may occur due to electromagnetic interference or other reasons. By setting a reasonable amplitude threshold, data points whose signal values ​​drastically change beyond the threshold within a short period of time can be considered outliers and replaced by linear interpolation of the preceding and following data.

[0039] After acquiring the real-time light intensity signal, this embodiment uses an integral model to calculate the average light loss rate within the detection interval. The mathematical model function is as follows: in, Indicates the average optical loss rate. This represents the real-time light intensity signal, where a and b represent the start and end points of the effective detection interval, respectively, and t represents the time variable, i.e., the time variable between the start and end points of the effective detection interval. The effective detection range is defined as one effective detection cycle of the chain harrow. Specifically, the starting point corresponds to the instant when one chain harrow completely leaves the light-transmitting slit and the photometric sensor begins to be exposed and detect crop stems. This instant is precisely determined by the moment when the output level signal of the chain harrow proximity sensor jumps from low (obstruction) to high (unobstructed). The ending point corresponds to the instant when the next chain harrow reaches and begins to obstruct the light-transmitting slit, and the photometric sensor is about to be shielded. This instant is precisely determined by the moment when the output level signal of the chain harrow proximity sensor jumps from high (unobstructed) to low (obstruction). The effective detection range is the length of one chain harrow operating cycle, precisely corresponding to the space between two adjacent harrow teeth. Within this range, the signals collected by the sensor fully reflect the degree of light obstruction by the crop stems trapped within the harrow window.

[0040] Compared to simply taking the average or peak value of the signal at several points within the interval, the integral model of this invention has significant advantages: Reflects total occlusion; integral The total light energy received by the sensor within the effective detection range (ba). Because It is the effective light intensity, and this integral value reflects the total amount of light energy that passes through the stem layer during this period.

[0041] It is insensitive to stem distribution; the distribution of crop stems in the conveying trough is random and uneven, and may be clump-like or sparse. Simple point sampling will produce huge fluctuations due to different stem distributions. In contrast, the integral value characterizes the total occlusion effect over the entire passing length, and can more stably reflect the total amount of stems passing through in that cycle. It is insensitive to the discrete distribution of stems and has strong anti-interference ability.

[0042] Normalization was performed; the denominator represents the baseline value of the total light energy that the sensor should have received within the same time interval under ideal, unobstructed conditions. Therefore, The actual physical meaning is: the percentage of total light energy loss due to crop stalk shading during one chain harrow cycle, i.e., the average light loss rate per unit length. This normalization process makes... By making it a dimensionless relative value, the impact of reference drift caused by factors such as light source aging and minute changes in sensor sensitivity on the final result is greatly reduced.

[0043] In this embodiment, the integration interval of the key parameter is not a preset fixed value, but dynamically determined. The values ​​of parameters a and b are determined in real time by the level transition signal output by the chain rake proximity sensor. As defined above, the starting point a is the moment when the level signal transitions from low to high, and the ending point b is the moment when it transitions from high to low again. This dynamically determined mechanism ensures that the integration calculation is always synchronized with the mechanical movement of the chain rake, so that each loss value strictly corresponds to a fixed physical space (the spacing of a rake window). This is crucial for achieving accurate monitoring of the feed rate, as it will change the signal over time. It is bound to the spatial dimension of the chain rake conveyor. Regardless of the chain rake speed, the conveying length corresponding to each detection interval is fixed, thus ensuring the accuracy of subsequent feed rate calculation.

[0044] Step 4: Based on the fitting curve of light loss rate and feed rate of the target crop at the standard chain harrow conveyor speed, determine the standard feed rate at the average light loss rate, and calculate the real-time feed rate by combining the real-time speed ratio of the chain harrow.

[0045] In this embodiment, the fitting curve between the light loss rate and the feed amount is established in the following way: During the experimental calibration phase, the combine harvester was controlled to operate at a preset, constant standard chain harrow conveyor speed. By varying the crop density of the test field, the combine harvester's travel speed, or the header height, a series of different feed rates covering the expected operating range were artificially created. For each stable feed rate, the following operations were performed simultaneously: Optical data acquisition and calculation: Record the voltage signal output by the photometric sensor group under this state, and calculate the average light loss rate (Loss) within one or more chain rake operation cycles according to the methods described in steps 1 to 3. To improve calibration accuracy, data can be continuously collected for a period of time (e.g., 30 seconds) under each stable state, and the average light loss rate of all effective detection intervals within this time period can be calculated as the final Loss value under this feed rate state.

[0046] Obtaining the actual feed rate of the target crop: The actual feed rate of the target crop is obtained using a weighing method. Specifically, during the same time period of the sustained steady state, the grain (or the corresponding material for stalks) after leaving the combine harvester's cleaning device is collected and weighed immediately, while the length of this time period is accurately recorded. The actual instantaneous feed rate (unit: kg / s) under this state can be calculated by dividing the mass of the collected material by the time. This value is calibrated as the actual feed rate of the target crop.

[0047] Preprocessing the acquired data points (average optical loss rate, actual feed rate) before curve fitting is crucial to ensure the accuracy and robustness of the fitted model. Preprocessing operations include, but are not limited to: Outlier removal: Statistical methods (such as the Laida criterion or box plot method) are used to identify and remove outliers that significantly deviate from the main data distribution area. These outliers may be caused by accidental factors during the experiment (such as sudden machine vibration, blockage by large foreign objects, etc.).

[0048] Data smoothing: When there are multiple feed rate measurements with similar loss values, the arithmetic mean of these measurements can be taken to form a more representative data point, thereby reducing the impact of random errors.

[0049] The obtained data points of actual feed rate and average light loss rate are organized to form a data point set. The least squares method is used for curve fitting to establish a quantitative functional relationship between the light loss rate and the feed rate. The quantitative functional relationship is a quadratic polynomial model, and its expression is: in, This is the standard feed rate at the standard chain rake conveyor speed. These are the crop-specific fitting coefficients for different target crops, obtained through curve fitting.

[0050] The crop-specific fitting coefficient These are the core parameters of the model, and their specific values ​​must be calculated using the least squares fitting algorithm based on a large amount of measured data through the aforementioned experimental calibration process. The method for determining these values ​​is as follows: The goal of the least squares method is to minimize the sum of the squares of the differences between the fitted values ​​and the true values ​​for all data points. The parameters that minimize this objective function can be easily solved using mathematical software or programming libraries (such as MATLAB, Python with NumPy / SciPy). The value of .

[0051] The quadratic polynomial model was chosen because the relationship between the feed rate and light loss rate of the target crop is usually not a simple linear one. Initially, the light loss rate increases approximately linearly with the increase in stem layer thickness; however, once the stem layer reaches a certain thickness, the increase in light loss rate per unit feed rate decreases due to the overlap and stacking of stems (i.e., a saturation trend emerges). The quadratic term effectively characterizes this nonlinear relationship. Compared to higher-order polynomials, the quadratic polynomial, while ensuring sufficient flexibility to fit nonlinear data, is less prone to overfitting, resulting in a more robust model with lower computational cost, making it suitable for implementation in embedded systems.

[0052] In this embodiment, a set of data points was used, and a quadratic polynomial model was used to fit the quantitative function relationship. The table below shows the specific data for some of the data points: Table 1: Data set of average optical loss rate and standard feed rate According to Table 1 above and Figure 2 The relationship between the standard feed rate and the average light loss rate can be clearly and intuitively observed. In this embodiment, the fitted quantitative function is specifically... The curve is based on a soybean harvester. A certain variety of soybeans is transported at an average speed of 1.5 m / s, with a feed rate ranging from 0.5 kg / s to 4 kg / s. The curve is fitted with a quadratic polynomial, and the mean absolute percentage error (MAPE) is 4.08%, with a coefficient of determination of 0.9895. Figure 3 This diagram illustrates the conventional residual distribution of the average light loss rate, visually showcasing the fitting accuracy and data reliability of the quadratic polynomial fitting model for light loss rate and feed amount established in this invention. The horizontal axis represents the average light loss rate, and the vertical axis represents the conventional residual (i.e., the difference between the actual feed amount and the model-fitted feed amount for each measured data point). The residual distribution shows that all residual data points are concentrated within the range of ±2, exhibiting a random and trendless distribution without systematic positive or negative bias clustering. This distribution result demonstrates that the quadratic polynomial fitting model used in this invention can effectively capture the nonlinear relationship between the target crop feed amount and the average light loss rate, achieving a good fitting effect and effectively reducing the impact of random errors on model accuracy. It also verifies the high reliability of the average light loss rate data calculated through the integral model, providing stable data support for the subsequent accurate calculation of real-time feed amount and further ensuring the measurement accuracy of the entire monitoring system.

[0053] As an alternative, the quantitative functional relationship can also be modeled in other forms, such as: Piecewise linear model: Divide the entire data range into several intervals, and fit a linear function to each interval. This method is computationally simpler.

[0054] Power function or exponential function model: If the shading characteristics of some crops exhibit stronger nonlinearity, this type of model can be considered.

[0055] Regardless of the model used, the specific values ​​of its coefficients must be determined through the aforementioned experimental calibration and fitting process.

[0056] In this embodiment, the The three crop-specific fitting coefficients are calibrated and stored separately for soybeans, rice, or wheat (e.g., stem diameter, density, flexibility, moisture content, etc.). For example, soybean stems are thick, and the shading area per unit mass may be completely different from that of slender, dense rice stems; therefore, different mapping relationships must be used. When the combine harvester switches harvested crop types, the corresponding fitting coefficients are called to update the quantitative function relationship. For example, in the combine harvester's control system, a crop type and fitting coefficient lookup table is preset. When the operator switches harvested crop types on the user interface, the system automatically calls the corresponding crop-specific fitting coefficients.

[0057] In real-time operation, the standard feed rate is calculated using the current average optical loss rate and the currently active fitting coefficients, according to the formula above. The specific formula for calculating the real-time feed rate, combined with the real-time transmission speed of the chain rake, is as follows: Where Q represents the real-time feed rate.

[0058] The calibration curve is at the standard speed ( The model is based on the principle that when the actual operating speed (V) of the chain rake changes, the stalk mass passing through the detection area per unit time will inevitably change. If the speed increases (P>1), the actual feed rate will increase proportionally under the same light loss rate; if the speed decreases (P<1), the actual feed rate will decrease proportionally. Therefore, introducing a speed ratio P for correction is a crucial step in converting the optical measurement model from the stalk mass per unit length to the stalk mass per unit time (i.e., feed rate), eliminating the impact of changes in operating speed on the monitoring results and ensuring the universality and accuracy of feed rate monitoring.

[0059] Step 5: Compare the real-time feed rate with the preset rated feed rate threshold, and generate a control signal based on the comparison result to adjust the travel speed or header height of the combine harvester, thereby achieving real-time control of the feed rate.

[0060] In this embodiment, a rated feed rate threshold is set, and the difference between the real-time feed rate and the rated feed rate threshold is calculated and used as the deviation, expressed by the formula: in, Indicates the deviation amount. This indicates the preset rated feed rate threshold.

[0061] Considering the fluctuations in sensor signals and the complexity of the field working environment, the instantaneous deviation is used directly. Making decisions can easily lead to frequent operation of actuators (such as hydraulic valves and motors), causing oscillations and affecting operational stability and machine lifespan. Therefore, before comparison, it is preferable to... Preprocessing is performed.

[0062] The preprocessing operation specifically involves: applying a first-order low-pass filter algorithm or a moving average filter algorithm to the initial deviation. Smoothing is performed to suppress high-frequency noise and transient disturbances.

[0063] For example: using a first-order low-pass filter algorithm: Where E(n) is the effective deviation after filtering in the current control cycle, and E(n-1) is the effective deviation in the previous control cycle. For filter coefficients ( ). The smaller the value, the stronger the filtering effect and the smoother the system response; The larger the value, the more sensitive it is to instantaneous changes. The specific value is determined through experiments under typical operating conditions, and is usually between 0.1 and 0.3, in order to achieve a balance between response speed and control stability.

[0064] For example, using a moving average filtering algorithm: Take the most recent N control cycles The values ​​are used to calculate the arithmetic mean of the values, which is then used as the current effective deviation. The value of N is usually between 5 and 10. If the value is too small, the filtering effect will be poor, and if it is too large, it will cause the system response to lag.

[0065] The effective deviation obtained after preprocessing will serve as the basis for subsequent logical judgments. The control cycle mentioned in this embodiment refers to a fixed time interval for the execution of the control system software; it is a preset, fixed value, such as 100 milliseconds or 50 milliseconds. It determines the frequency at which the control system calculates the deviation, executes the control logic, and outputs the control signal.

[0066] If the deviation is greater than the preset first positive threshold, it is determined to be overfeeding, and a control signal is generated to reduce the real-time feeding amount. If the deviation is less than the preset first negative threshold, it is determined that the feed is insufficient, and a control signal is generated to increase the real-time feed amount. If the deviation is within the preset range of the first negative threshold and the first positive threshold, it is determined that the feed amount is appropriate, and a control signal is generated to maintain the current operating state.

[0067] The specific values ​​of the first positive threshold and the first negative threshold should be determined comprehensively based on the model of the combine harvester, the type of crop to be harvested, and the required control precision.

[0068] For example, based on machine performance, the threshold range should be smaller than the maximum feed rate fluctuation range that the combine harvester can withstand. For instance, for a harvester with a rated feed rate of 10 kg / s, the initial setting of the absolute values ​​of the first positive threshold and the first negative threshold can be considered as 5%-15% of its rated value, i.e., 0.5 kg / s to 1.5 kg / s. This range ensures timely intervention in the event of actual overload or starvation, while avoiding unnecessary adjustments within the normal fluctuation range.

[0069] For example, based on the control objective, if extremely high operational stability is desired, the threshold range can be appropriately narrowed to make the control system more sensitive; if priority is given to the lifespan of actuators such as hydraulic systems and fuel economy, the threshold range can be appropriately widened to reduce the frequency of adjustments. Therefore, the final values ​​of the first positive threshold and the first negative threshold are optimized through repeated trials and calibrations in the target crop field, and their rationality lies in achieving the best balance between control accuracy and actuator lifespan.

[0070] The control signal is configured to directly control the actuator of the combine harvester, specifically as follows: the control signal for reducing the real-time feed rate is configured to perform at least one of the following operations, namely, reducing the travel speed of the combine harvester or increasing the header height.

[0071] Reducing the travel speed of the combine harvester is the most direct and effective way to reduce its load. The control signal can be an analog voltage or a PWM wave, sent to the harvester's hydraulic continuously variable transmission (CVT) system, instructing it to reduce its forward speed in preset steps (e.g., 0.1 km / h each time) or proportionally based on the magnitude of the effective deviation. The reduction in travel speed can be directly proportional to the magnitude of the effective deviation or a segmented PID control strategy can be used to achieve smoother control.

[0072] Raising the header height indirectly reduces the feed rate by decreasing the amount of crop the header can collect. A control signal is sent to the hydraulic cylinders that regulate the header's height. When performing this operation, it's crucial to ensure the header height remains within agronomically permissible limits to prevent crop drop or excessive stubble height. Therefore, the system can integrate a header height limit sensor to ensure the control signal operates within a safe range.

[0073] The control signal used to increase the real-time feed rate is configured to perform at least one of the following operations: increasing the travel speed of the combine harvester or decreasing the header height. Increasing the travel speed of the combine harvester improves operational efficiency. The control logic is similar to that for deceleration, increasing the speed in steps or proportionally. Decreasing the header height allows the header to catch more crop.

[0074] To optimize the control effect, the generation of the control signal can be further optimized using a proportional-integral-derivative (PID) control algorithm. Specifically, the system uses the effective deviation as the input to the PID controller to calculate the control output. The values ​​of the proportional, integral, and derivative terms are determined using engineering tuning methods, such as the Ziegler-Nichols method, and fine-tuned in actual field operations to ensure that the control system responds quickly, has small overshoot, and is stable.

[0075] Furthermore, when simultaneously selecting strategies to adjust both travel speed and header height, priorities can be set. For example, travel speed can be adjusted as the primary means, and header height adjustment can be initiated as a secondary means when the speed adjustment has reached its limit (such as the minimum safe speed or the maximum design speed) and the feed rate has not yet returned to the appropriate range.

[0076] Please see Figure 4 The present invention also provides a feed rate monitoring device for light loss detection, which is used to implement the above-mentioned feed rate monitoring method for light loss detection, comprising: The system calibration and self-calibration module is used to turn on the light source set at the bottom of the conveying trough when there are no stalks passing through the conveying trough of the combine harvester. The light shines through the light-transmitting slit at the bottom of the conveying trough onto the photometric sensor group directly above, and measures and stores the voltage value output by the photometric sensor group at this time as the reference light intensity value. The signal acquisition and preprocessing module is used to acquire the real-time light intensity signal output by the photometric sensor group when the crop stalk passes through the light-transmitting slit during the harvesting operation of the combine harvester. It also acquires the real-time transmission speed of the chain rake through the chain rake proximity sensor and calculates the real-time speed ratio of the chain rake by combining it with the preset standard chain rake transmission speed. The light loss rate analysis module is used to calculate the average light loss rate of crop stems per unit length based on the reference light intensity value and the real-time light intensity signal. The feed rate mapping and correction module is used to determine the standard feed rate under the average light loss rate based on the fitting curve of the light loss rate of the target crop and the feed rate under the standard chain harrow transmission speed, and to calculate the real-time feed rate by combining the real-time speed ratio of the chain harrow. The decision and control module is used to compare the real-time feed rate with the preset rated feed rate threshold, and generate control signals based on the comparison results to adjust the travel speed or header height of the combine harvester, thereby achieving real-time control of the feed rate.

[0077] 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.

[0078] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0079] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0080] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes 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.

Claims

1. A method for monitoring feed rate with light loss detection, characterized in that, The specific steps include: Step 1: With no stalks passing through the conveying trough of the combine harvester, turn on the light source set at the bottom of the conveying trough. The light shines through the light-transmitting slit at the bottom of the conveying trough onto the photometric sensor group directly above. Measure and store the voltage value output by the photometric sensor group at this time as the reference light intensity value. Step 2: During the harvesting of the target crop by the combine harvester, the real-time light intensity signal output by the photometric sensor group when the crop stalk passes through the light-transmitting slit is collected in real time, and the real-time transmission speed of the chain rake is obtained by the chain rake proximity sensor, so as to calculate the chain rake real-time speed ratio by combining it with the preset standard chain rake transmission speed. Step 3: Based on the reference light intensity value and the real-time light intensity signal, calculate the average light loss rate of the crop stem per unit length. Step 4: Based on the fitting curve of light loss rate of target crop and feed rate at standard chain harrow transmission speed, determine the standard feed rate at average light loss rate, and calculate the real-time feed rate by combining the real-time speed ratio of chain harrow. Step 5: Compare the real-time feed rate with the preset rated feed rate threshold, and generate a control signal based on the comparison result to adjust the travel speed or header height of the combine harvester, thereby achieving real-time control of the feed rate.

2. The method for monitoring feed volume for detecting light loss according to claim 1, characterized in that, The light source, the light-transmitting slit, and the photometric sensor group constitute a closed optical sensing unit. The light-transmitting slit is a 1.5 cm wide through-slot located at the bottom of the conveying trough, used to constrain the light emitted by the light source into a narrow light band. The photometric sensor group consists of multiple photometric sensors arranged uniformly above the light-transmitting slit to receive the light passing through it. The photometric sensor includes a cosine correction plate, a silicon photodiode, and a signal conditioning circuit arranged sequentially along the optical path. The cosine correction plate is used to correct the incident light angle according to Lambert's cosine law. The signal conditioning circuit is a transimpedance amplifier circuit used to convert the photocurrent signal generated by the silicon photodiode into a voltage signal output.

3. The feeding amount monitoring method for light loss detection according to claim 1, characterized in that, Obtaining the real-time light intensity signal, real-time transmission speed, and the ratio of the chain rake's real-time speed specifically includes: The photometric sensor array continuously acquires light intensity signals at a sampling frequency of not less than 1kHz and outputs corresponding voltage values; simultaneously, the chain rake proximity sensor monitors the passing status of the chain rake in real time and outputs corresponding level signals. When the chain rake proximity sensor outputs a low-level signal indicating that the chain rake is approaching, it is determined that the light-transmitting slit is completely blocked by the chain rake at this time. The light intensity signal collected by the photometric sensor is an invalid signal, and the light intensity signal in this period is marked as invalid and not used. The moment when the output level signal of the chain rake proximity sensor changes from low to high is taken as the start of the effective detection interval, and the moment when the output level signal changes from high to low is taken as the end of the effective detection interval. The real-time light intensity signal is taken from the photometric sensor data within the effective detection interval. The chain rake proximity sensor monitors the time interval between the transition from a low-level signal to a high-level signal between two adjacent rake teeth, and calculates the ratio of the fixed distance between the two adjacent rake teeth to the time interval. This ratio is the real-time transmission speed of the chain rake. The real-time speed ratio of the chain rake is the ratio of the real-time transmission speed to the preset standard chain rake transmission speed.

4. The method for monitoring feed amount for light loss detection according to claim 2, characterized in that, The calculation of the average optical loss rate specifically includes: The average optical loss rate within the effective detection range is calculated using an integral model, and its mathematical model function is as follows: in, Indicates the average optical loss rate. This represents the real-time light intensity signal, where a and b represent the start and end points of the effective detection interval, respectively, and t represents the time variable. The effective detection range is defined as one effective detection cycle of the chain rake. Specifically, the starting point corresponds to the instant when a chain rake completely leaves the light-transmitting slit and the photometric sensor begins to detect the crop stem; the ending point corresponds to the instant when the next chain rake reaches and begins to block the light-transmitting slit and the photometric sensor is about to be shielded; the effective detection range is the length of one chain rake operation cycle.

5. The method for monitoring feed amount for light loss detection according to claim 4, characterized in that, The fitting curve between the light loss rate and the feed rate is established in the following way: During the experimental calibration phase, the combine harvester was controlled to run at a constant speed at the standard chain harrow transmission speed, and data points were obtained under multiple different feeding conditions by changing the feed amount of the target crop. For each data point, record the actual feed amount obtained by weighing and calculate the corresponding average optical loss rate. The obtained data points of actual feed rate and average light loss rate are organized to form a data point set. The least squares method is used for curve fitting to establish a quantitative functional relationship between the light loss rate and the feed rate. The quantitative functional relationship is a quadratic polynomial model, and its expression is: in, This is the standard feed rate at the standard chain rake conveyor speed. These are the crop-specific fitting coefficients for different target crops, obtained through curve fitting. Based on the real-time conveying speed of the chain rake, the specific formula for calculating the real-time feed rate is as follows: Where Q represents the real-time feed volume, This represents the real-time speed ratio of the chain rake.

6. The method for monitoring feed volume for detecting light loss according to claim 5, characterized in that, The The three crop-specific fitting coefficients are calibrated and stored separately for soybeans, rice, or wheat. When the combine harvester switches to harvest a different crop, the corresponding fitting coefficients are called to update the quantitative function relationship.

7. The method for monitoring feed amount for light loss detection according to claim 5, characterized in that, Set a rated feed rate threshold, calculate the difference between the real-time feed rate and the rated feed rate threshold, and use it as the deviation. If the deviation is greater than the preset first positive threshold, it is determined to be overfeeding, and a control signal is generated to reduce the real-time feeding amount. If the deviation is less than the preset first negative threshold, it is determined that the feed is insufficient, and a control signal is generated to increase the real-time feed amount. If the deviation is within the preset range of the first negative threshold and the first positive threshold, it is determined that the feed amount is appropriate, and a control signal is generated to maintain the current working state. The control signal is configured to directly control the actuator of the combine harvester, specifically as follows: the control signal for reducing the real-time feed rate is configured to perform at least one of the following operations, namely, reducing the travel speed of the combine harvester or increasing the header height; the control signal for increasing the real-time feed rate is configured to perform at least one of the following operations, namely, increasing the travel speed of the combine harvester or reducing the header height.

8. A feed rate monitoring device for detecting light loss, characterized in that, The feed rate monitoring device for optical loss detection is used to implement the feed rate monitoring method for optical loss detection according to any one of claims 1-7, comprising: The system calibration and self-calibration module is used to turn on the light source set at the bottom of the conveying trough when there are no stalks passing through the conveying trough of the combine harvester. The light shines through the light-transmitting slit at the bottom of the conveying trough onto the photometric sensor group directly above, and measures and stores the voltage value output by the photometric sensor group at this time as the reference light intensity value. The signal acquisition and preprocessing module is used to acquire the real-time light intensity signal output by the photometric sensor group when the crop stalk passes through the light-transmitting slit during the harvesting operation of the combine harvester. It also acquires the real-time transmission speed of the chain rake through the chain rake proximity sensor and calculates the real-time speed ratio of the chain rake by combining it with the preset standard chain rake transmission speed. The light loss rate analysis module is used to calculate the average light loss rate of crop stems per unit length based on the reference light intensity value and the real-time light intensity signal. The feed rate mapping and correction module is used to determine the standard feed rate under the average light loss rate based on the fitting curve of the light loss rate of the target crop and the feed rate under the standard chain harrow transmission speed, and to calculate the real-time feed rate by combining the real-time speed ratio of the chain harrow. The decision and control module is used to compare the real-time feed rate with the preset rated feed rate threshold, and generate control signals based on the comparison results to adjust the travel speed or header height of the combine harvester, thereby achieving real-time control of the feed rate.

Images