A method for identifying parameters of a mechanism model of a combine harvester threshing and cleaning operation quality
By constructing a quality mechanism model for combine harvester threshing and cleaning operations and adopting a multi-strategy improved whale optimization algorithm, the problems of missing quality models and low parameter identification accuracy for combine harvester threshing and cleaning operations were solved, achieving high-precision operation quality prediction and system optimization.
Patent Information
- Application Number
- CN202611375546.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-04
- Publication Date
- 2026-11-03
AI Technical Summary
In the existing technology, there is a lack of a mechanism model for the quality of threshing and cleaning operations in combine harvesters. Traditional parameter identification methods suffer from low identification accuracy, slow convergence speed, and poor stability, making it difficult to accurately describe the intrinsic relationship between operation parameters and operation quality.
A quality mechanism model for threshing and cleaning operations was constructed, and a multi-strategy improved whale optimization algorithm was used for parameter identification. Tent chaotic mapping, nonlinear adaptive convergence factor, nonlinear inertial weight, Lévy flight mechanism and sparrow search algorithm were integrated, and model parameters were identified in combination with measured data.
It achieves a high-precision description of the relationship between operation parameters and operation quality, improves the model prediction accuracy and stability, and provides a foundation for performance optimization and intelligent control of the combine harvester threshing and cleaning system.
Smart Images
Figure CN123030908A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent optimization and parameter identification technology for agricultural machinery, specifically to a method for identifying parameters of a quality mechanism model for threshing and cleaning operations of a combine harvester. Background Technology
[0002] The threshing and cleaning system of a combine harvester is a key working component that affects harvest quality and loss levels. Its performance directly relates to indicators such as grain breakage rate, entrainment loss rate, impurity content, and cleaning loss rate.
[0003] Currently, research on the quality of threshing and cleaning operations is still in the stage of detecting operating parameters and operating quality. The threshing and cleaning process involves complex mechanical behaviors such as collisions between crop particles, vibration and transport of the screen surface, and coupling and separation of gas and solid phases. The operating parameter adjustment method without a mechanistic model is difficult to adapt to complex working conditions.
[0004] For modeling the operation process of threshing and cleaning systems, traditional polynomial response surface analysis is the most commonly used method to optimize the performance of combine harvester threshing and cleaning systems. Although data-driven modeling methods can describe the nonlinear and strong coupling characteristics between input and output variables, they lack an essential description of the quality of threshing and cleaning operations.
[0005] Model parameter identification can employ the least squares method and intelligent optimization methods such as genetic algorithms and particle swarm optimization. However, the threshing and cleaning process is characterized by multi-parameter coupling and nonlinearity. Traditional parameter identification methods are prone to getting trapped in local optima, resulting in problems such as low identification accuracy, slow convergence speed, and poor stability. They are difficult to accurately describe the intrinsic relationship between operation parameters and operation quality.
[0006] Therefore, it is essential to design a parameter identification method for the quality mechanism model of combine harvester threshing and cleaning operations. Summary of the Invention
[0007] The purpose of this invention is to provide a method for identifying parameters of a combine harvester threshing and cleaning operation quality mechanism model. While establishing and identifying the operation quality mechanism model of a grain combine harvester, this method can reveal the influence mechanism of the operation parameters of the combine harvester threshing and cleaning system on the operation quality.
[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester, comprising: Step 1: Construct a threshing and cleaning operation quality mechanism model that characterizes the mapping relationship between operation parameters and operation quality; the operation parameters include at least the feeding speed, drum speed, threshing gap, blower speed and screen opening; the operation quality includes at least the grain breakage rate, entrainment loss rate, impurity content and cleaning loss rate; Step 2: Obtain measured data of operating parameters and corresponding measured data of operating quality of the combine harvester under various combinations of operating parameters to form a dataset; Step 3: Construct a multi-strategy improved whale optimization algorithm and use the dataset to identify unknown parameters in the mechanistic model to determine the optimal model parameters, minimizing the error between the job quality predicted by the mechanistic model based on the job parameters and the measured job quality data; wherein, the multi-strategy improved whale optimization algorithm integrates the Tent chaotic mapping initialization population strategy, the nonlinear adaptive convergence factor strategy, the nonlinear inertia weight strategy, the Lévy flight mechanism, the sparrow search algorithm development mechanism, and the Cauchy-Gaussian mutation strategy; Step 4: Validate the prediction accuracy of the mechanistic model with identified parameters using validation data independent of the dataset.
[0009] Furthermore, in step 1, a mechanism model for the quality of threshing and cleaning operations is constructed to characterize the mapping relationship between operational parameters and operational quality. Specifically: Based on the assumption of steady-state material flow mass conservation, the threshing and cleaning system of the combine harvester is regarded as a whole, and the mechanism modeling is carried out for the threshing process and the cleaning process respectively. In the modeling of the threshing process, a breakage rate mechanism model for calculating the grain breakage rate and an entrainment loss rate mechanism model for calculating the entrainment loss rate are constructed. In the modeling of the cleaning process, a material looseness coefficient model for the screen surface is constructed to characterize the looseness of the material. Based on the looseness coefficient model, a mechanism model for calculating the impurity rate and a mechanism model for calculating the cleaning loss rate are constructed respectively.
[0010] Furthermore, a breakage rate mechanism model is constructed, specifically as follows: A differential equation for grain crushing quality was established by deriving and relating the material layer volume fraction and the average number of collisions. Based on the relationship between threshing time and the axial velocity of the material flow, the differential equation is solved to obtain the calculation expression for the grain breakage rate. The calculation expression characterizes the grain breakage rate as a power function including the feeding speed, drum speed and threshing gap, and a comprehensive influence coefficient composed of material properties and structural parameters.
[0011] Furthermore, a mechanism model for the entrainment loss rate is constructed, specifically as follows: A differential equation for the mass of grain entrainment loss is established, taking the condition that the axial velocity of the material is greater than the radial velocity. The differential equation is solved based on the threshing time to obtain the calculation expression for the grain entrainment loss rate. The calculation expression characterizes the grain entrainment loss rate as a power function including the feeding speed, drum speed and threshing gap, as well as a comprehensive influence coefficient composed of material properties and structural parameters.
[0012] Furthermore, a model for the bulk density coefficient of the material on the screen surface is constructed, specifically as follows: Construct calculation expressions for ideal material layer thickness, material layer thickness under vibration, and material layer thickness under airflow. Based on the material layer thickness under vibration and airflow, an expression for calculating the material layer thickness under the coupled action of vibration and airflow is constructed. The ratio of the material layer thickness under coupling effect to the ideal material layer thickness is defined as the material looseness coefficient of the screen surface, in order to correlate the influence of fan speed and screen opening on the looseness of the material.
[0013] Furthermore, a mechanism model for impurity rate and a mechanism model for cleaning loss rate are constructed, specifically as follows: The calculation expression for the material looseness coefficient on the screen surface is substituted into the differential equations for impurity mass and grain loss mass established based on the concept of mass flow, respectively. By solving the differential equations and substituting variables based on the cleaning time, the calculation expressions for the impurity rate and the cleaning loss rate are obtained. The impurity rate calculation expression characterizes the impurity rate as a function of the feed rate, blower speed, screen opening, and the impurity rate influence coefficient. The cleaning loss rate calculation expression characterizes the cleaning loss rate as a function of the feed rate, blower speed, screen opening, and the cleaning loss rate influence coefficient.
[0014] Furthermore, in step 2, the measured data of the operating parameters and the corresponding measured data of the operating quality of the combine harvester under various combinations of operating parameters are obtained, specifically as follows: Amplitude-modulated pseudo-random binary sequence signal is used as excitation signal to automatically adjust the operating parameters of the combine harvester over a wide range and at multiple sample points. The system collects real-time operating parameters and corresponding real-time operating quality data of the combine harvester under the excitation signal.
[0015] Furthermore, the breakage rate and impurity rate in the real-time operation quality data are obtained through an image recognition method; the image recognition method includes: obtaining samples through a sampling device, acquiring sample images using an industrial camera, distinguishing between whole grains, broken grains and impurities by analyzing the pixel area features of particles in the image, and calling a calculation model to output the breakage rate and impurity rate; The entrainment loss rate and cleaning loss rate in the real-time operation quality data are obtained through piezoelectric sensors. The piezoelectric sensors detect and analyze the amplitude, duration, and frequency characteristics of the piezoelectric signals generated by the material impacting the sensitive plate to distinguish between grains and impurities, and calculate the loss rate based on statistical data and operation parameters.
[0016] Furthermore, in step 3, a multi-strategy improved whale optimization algorithm is constructed, specifically as follows: An improved Tent chaotic mapping method was used to initialize the whale population location to enhance the diversity and uniformity of the initial population. A nonlinear adaptive convergence factor strategy is adopted to balance the algorithm's global exploration and local exploitation capabilities; Nonlinear inertial weights and Lévy flight mechanism are introduced during the position update process to enhance the algorithm's ability to escape local optima. The development mechanism of the sparrow search algorithm is integrated, and the position of poor individuals is updated by a specific perturbation probability to improve the overall optimization ability of the population. A Cauchy-Gaussian mutation strategy is applied to the current optimal solution to improve the accuracy of the solution and prevent premature convergence of the algorithm.
[0017] Furthermore, in step 3, the unknown parameters in the mechanistic model are identified using the dataset to determine the optimal model parameters, specifically as follows: The model parameters to be identified are used as optimization variables, and a fitness function is constructed based on the error between the job quality predicted by the mechanism model and the measured job quality data. The whale optimization algorithm is improved by using multiple strategies for iterative optimization. When the fitness function value is minimized or the termination condition is met, the current model parameters are output as the optimal model parameters.
[0018] In summary, the beneficial technical effects of the present invention are as follows: 1. Based on the analysis of the threshing and cleaning operation process, this invention establishes a quality mechanism model for the threshing and cleaning operation, thereby realizing a mechanistic description of the relationship between operation parameters and operation quality; 2. This invention utilizes a multi-strategy improved whale optimization algorithm to enhance population diversity, improve global search capability and convergence speed, achieve high-precision identification of mechanistic model parameters, and improve model prediction accuracy. 3. This invention can provide a theoretical basis and model support for parameter optimization, model predictive control, and intelligent operation control of the threshing and cleaning system of combine harvesters; 4. The mechanism model and parameter identification method constructed in this invention have good versatility and can be extended to other fields of parameter identification and optimization of complex agricultural equipment systems.
[0019] In summary, the technical solution of this invention solves the problems of missing quality mechanism models for threshing and cleaning operations in existing combine harvesters, low accuracy of model parameter identification, and the tendency of identification algorithms to get trapped in local optima. It provides a reliable model foundation for performance prediction, parameter optimization, and intelligent control of combine harvester threshing and cleaning systems. Compared with existing parameter identification methods, this invention has advantages such as high identification accuracy, good stability, strong adaptability, and high versatility, and can be widely applied in the fields of complex system modeling, parameter identification, and optimized control of agricultural equipment. Attached Figure Description
[0020] Figure 1 This is a schematic diagram of a steady-state material flow model; Figure 2 This is a schematic diagram of the cleaning process under the coupling effect of airflow and vibration. Figure 3 This is a schematic diagram showing the adjustment of the fish scale screen opening. Figure 4 A flowchart illustrating the multi-strategy improvement of the whale optimization algorithm; Figure 5 Schematic diagram of the sensor for detecting breakage rate and impurity content; Figure 6 A schematic diagram of the installation of the entrainment loss rate detection sensor; Figure 7 Schematic diagram of the installation of the cleaning loss rate detection sensor; Figure 8 This is a schematic diagram of an electric adjustment device for the threshing gap; Figure 9 Schematic diagram of the electric adjustment device for screen opening; Figure 10 A schematic diagram illustrating the fitness change curve for the parameters; Figure 11 A schematic diagram comparing the parameter identification output of the mechanism model with the actual output; Figure 12 Schematic diagram of the verification results of mechanism model parameter identification; Figure 13 This is a schematic diagram of the method flow of the present invention. Detailed Implementation
[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.
[0022] like Figure 13As shown, this invention provides a method for identifying parameters of a quality mechanism model for threshing and cleaning operations in combine harvesters, including: Step 1: Construct a threshing and cleaning operation quality mechanism model that characterizes the mapping relationship between operation parameters and operation quality; the operation parameters include at least the feeding speed, drum speed, threshing gap, blower speed and screen opening; the operation quality includes at least the grain breakage rate, entrainment loss rate, impurity content and cleaning loss rate; Step 2: Obtain measured data of operating parameters and corresponding measured data of operating quality of the combine harvester under various combinations of operating parameters to form a dataset; Step 3: Construct a multi-strategy improved whale optimization algorithm and use the dataset to identify unknown parameters in the mechanistic model to determine the optimal model parameters, minimizing the error between the job quality predicted by the mechanistic model based on the job parameters and the measured job quality data; wherein, the multi-strategy improved whale optimization algorithm integrates the Tent chaotic mapping initialization population strategy, the nonlinear adaptive convergence factor strategy, the nonlinear inertia weight strategy, the Lévy flight mechanism, the sparrow search algorithm development mechanism, and the Cauchy-Gaussian mutation strategy; Step 4: Validate the prediction accuracy of the mechanistic model with identified parameters using validation data independent of the dataset.
[0023] Next, we will elaborate on the above method in detail with specific parameter settings and formulas.
[0024] In step 1, a mechanism model for the threshing and cleaning operation quality is constructed to characterize the mapping relationship between operation parameters and operation quality. The operation parameters include at least the feeding speed, drum speed, threshing gap, blower speed, and screen opening. The operation quality includes at least the grain breakage rate, entrainment loss rate, impurity content, and cleaning loss rate, specifically including: S1. Assumptions are made about the operation process of the combine harvester based on the steady-state material flow quality model; S2. Derive the expression for calculating the grain breakage rate in the threshing system; S3. Derive the expression for calculating the grain entrainment loss rate in the threshing system; S4. Derive the expression for calculating the bulk density coefficient of the screen material in the cleaning system; S5. Derive the expression for calculating the impurity content of grains in the cleaning system; S6. Derive the expression for calculating the cleaning loss rate in the cleaning system; S1-S6 will be explained in detail respectively: S1 describes the operation process of the threshing and cleaning system, and constructs a steady-state material flow model (such as...). Figure 1 (As shown), feeding amount This can be achieved by adjusting crop density and operating speed; the threshed mixture after threshing is divided into grain weight. Non-grain impurity quality The mass discharged through the end of the threshing drum is divided into the mass of grains lost due to entrainment. Incomplete removal of waste results in quality loss. and stem quality Entrainment loss mass This indicates that the grains were not separated sufficiently and were mixed in with the stalk layer, resulting in losses that were discharged from the combine harvester.
[0025] The cleaning system removes the grains from the mixture by weight. and impurity quality Effective separation is achieved, thereby reducing the quality loss during grain cleaning. Residual quality And the quality of the grain entering the granary. and impurity quality .
[0026] S2 derives the differential equation for grain breakage and the expression for calculating the breakage rate. After the crop enters the threshing system, the feeding rate remains stable, and a stable material layer is formed between the concave plate and the threshing element. The volume fraction of the material layer... With the radius of the roller R and threshing gap The relevant calculation formula is as follows: ; in, It is the effective radius of the threshing element. R It is the roller radius and It is the threshing gap. .
[0027] The average number of collisions of the material flow within the threshing system is related to parameters such as the number of threshing elements, drum speed, feed rate, and material volume fraction. The calculation formula is as follows: ; in, N This indicates the drum rotation speed, in r / min; Z represents the number of collisions between the material flow and the threshing device; Z represents the number of threshing elements.
[0028] The threshing element strikes the material flow, causing the crop to thresh; this process is an energy accumulation effect under the impact force. The impact velocity of the threshing element on the material flow... Related to drum speed: ; The energy experienced by the material during threshing is obtained based on the impact velocity. This energy is calculated using the average number of impacts, as shown in the following formula: ; in, Represents the energy of a single collision, in J; The average collision energy is expressed in J. based on The related calculation formula is derived as follows: Complete calculation expression: ; The energy applied to the material by the threshing element is also transferred to the grains. Assuming the energy threshold for critical grain breakage is... The critical condition for grain breakage is: Since collision energy is difficult to measure, the critical condition for grain breakage is the critical rotational speed threshold, calculated using the following formula: ; in, This indicates the critical rotational speed for grain crushing, in r / min.
[0029] The key operating parameters in the grain threshing process are mainly the feed rate, drum speed, and threshing gap. Based on these key operating parameters and the mass flow characteristics during threshing, a differential equation for grain breakage in the threshing system is established: ; in, Indicates the quality of intact threshed grains. This indicates the mass of the broken grains. Indicates the grain breakage coefficient. Indicates the coefficient of grain breakage. , and These represent the power functions that affect the feeding speed, drum speed, and threshing gap during grain crushing, respectively.
[0030] Solving the differential equation for grain breakage by integration, given the initial conditions: hour, ,but The quality of the crushed grains obtained from the threshing system is as follows: ; Breakage rate The calculation expression is: ; The time from soybean crop feeding to complete threshing and separation is measured by... express ; in, L The length of the roller is in meters (m). The axial velocity of the material flow is expressed in m / s.
[0031] The axial velocity of the material flow is mainly generated by the screw conveyor of the drum, and is primarily related to the drum rotation speed, guide plate angle, friction coefficient, and feed rate. Based on the circumferential velocity generated by the rotation of the threshing drum, the axial velocity of the material flow is... ; in, For the angle of the deflector, ; For material conveying efficiency, Bulk density of the material, kg / m³ 3 .
[0032] Based on the drum length and the axial speed of the material, the threshing time can be derived. : ; When calculating grain breakage, use replace The final expression for calculating the breakage rate can be obtained as follows: ; in, This indicates the influence coefficient of material properties and structural parameters on the breakage rate. These parameters include the grain breakage coefficient, drum length, material bulk density, drum diameter, and guide vane angle.
[0033] In section S3, the differential equation for grain entrainment loss and the expression for calculating the entrainment loss rate are derived. The condition for grain entrainment loss is that the axial velocity is greater than the radial velocity. The condition for entrainment loss to occur is: ; when At this time, the threshing system produces entrainment loss, and the differential equation for grain entrainment loss is: ; in, This indicates the mass of grains lost due to entrainment. Indicates the weight of the threshed grains. Indicates the grain entrainment loss coefficient. Indicates the coefficient representing the degree of grain entrainment loss. , and These represent the power functions that affect the grain entrainment loss, namely, the feed rate, the drum speed, and the threshing gap.
[0034] Solving the differential equation, we get: ; Entrainment loss rate calculation formula: ; During the threshing process, the time required for the grains to be threshed is... ,use represent Then, the final formula for calculating the entrainment loss rate mechanism model is obtained: ; in, This indicates the influence coefficient of material properties and structural parameters on the entrainment loss rate. These parameters include the grain breakage coefficient, drum length, material bulk density, drum diameter, and guide plate angle.
[0035] In S4, the formula for the bulk density coefficient of the screen material is derived. The essence of cleaning is the separation process of grains and MOGs under the combined action of a vibrating screen surface and an airflow field. The cleaning process under the coupled action of airflow and vibration is as follows: Figure 2 As shown. Under ideal conditions of no airflow and no fish-scale screen vibration, the discharged mixture forms an ideal material layer on the screen surface, with a material layer thickness of... The calculation formula is: ; in, The ideal material layer thickness is expressed in meters (m). The width of the sieve surface is expressed in meters (m). The length of the sieve surface is expressed in meters (m). This indicates the total mass of the grain and MOG mixture; This indicates the density of the mixture of grains and MOG.
[0036] Under the action of a vibrating screen, the material on the screen surface is thrown and becomes loose, thus giving the grains a greater chance to pass through the material layer and contact the screen surface. Therefore, the throwing intensity of the vibrating screen needs to be greater than 1. The thickness of the material layer under the vibration of the vibrating screen... The calculation formula is: ; in, This represents the coefficient indicating the influence of the screen throwing intensity on the material layer. Indicates the direction angle of vibration; Indicates the inclination angle of the sieve surface; Indicates the vibration intensity of the fish scale screen; This indicates the vibration amplitude of the fish scale screen.
[0037] The motion state of the mixture on the screen surface is also related to the airflow intensity generated by the blower and the volumetric flow rate at the blower outlet. and the airflow velocity on the screen surface The calculation formula is: ; in, Indicates the area of the air outlet, in meters. 2 ; Indicates the airflow velocity at the air outlet, in m / s; Indicates the characteristic parameters of the fan; This indicates the fan speed, in r / min; This indicates the open area ratio of the sieve surface; This represents the attenuation coefficient of the airflow reaching the screen surface.
[0038] Assuming the expansion state of the material layer due to airflow is indirectly calculated using the airflow velocity on the screen surface, the thickness of the material layer under the coupled action of vibration and airflow is... The calculation formula is: ; in, This represents the coupling coefficient between vibration and airflow; This represents the material layer coefficient under the action of airflow on the screen surface.
[0039] Based on the ideal material layer and the coupling effect of vibration and airflow, the material layer thickness can be determined by the degree of looseness. This describes the loose, fluidized state. Material bulkiness is the result of the combined effect of the airflow velocity on the screen surface and the vibrating screen; its essence is the porosity formed by the particle size difference between materials. The formula for calculating material bulkiness is: ; in, This represents the bulk density coefficient of the material on the screen surface.
[0040] After the particle group on the screen surface stratifies and migrates, the stratified material layer is then screened. The screening characteristics of grains and stems are related to the opening ratio of the fish-scale screen, and the opening of the fish-scale screen is adjusted as follows: Figure 3 As shown. When the fish scale screen is open, the opening degree of the fish scale screen. The calculation formula is: ; in, Indicates the inclination angle of the fish scale sieve, in degrees; The length of a single fish scale sieve is indicated in mm; When the inclination angle of the fish scale screen is At this time, the opening area of the fish scale sieve is The open area ratio of the fish scale screen Calculation formula: ; in, Indicates the number of arrangements along the length of the fish scale sieve; This indicates the number of fish scale screens in the width direction; This indicates the width of the entire fish scale sieve.
[0041] Formula for calculating looseness: ; Looseness is related to the characteristics of the blower, the structure of the fish scale screen, the vibration characteristics of the fish scale screen, and the properties of the material. Increasing the blower speed is beneficial to improving looseness, while increasing the opening of the fish scale screen will weaken looseness.
[0042] In S5, the formula for calculating the impurity content of grains is derived. During the cleaning process, when the airflow velocity on the screen surface exceeds the grain suspension velocity, the grains are easily blown away by the airflow, resulting in cleaning losses. The differential equation for grain impurity content can be expressed as: ; in, Indicates the quality of the grain warehouse MOG. Indicates the quality of MOGs entering the cleaning system. This indicates the generation of MOG coefficients. , and These represent the power functions that affect the feeding rate, fan speed, and screen opening when MOG enters the grain silo.
[0043] In S5, substituting the expression for calculating looseness into the differential equation for grain impurity, we get: ; in, This represents the impurity coefficient of the grains, which is related to the structural parameters of the cleaning system and crop attributes.
[0044] In S5, the differential equation for impurity content in the grains during the cleaning process is solved to obtain the formula for calculating the impurity mass: ; In S5, the impurity content The calculation formula is: ; In S5, the cleaning time of the desorbed mixture in the cleaning system. Related to the airflow velocity and screen length, the calculation is expressed as follows: ; In S5, using replace The expression for calculating the impurity rate is obtained as follows: ; in, This represents the impurity content influence coefficient, which is related to the structural parameters of the fish scale screen and the fan.
[0045] In S6, substituting the expression for calculating looseness into the differential equation for grain loss, we get: ; In S6, the formula for calculating the mass loss during grain cleaning is: ; In S6, the cleaning loss rate The calculation formula is ; In S6, using replace The formula for calculating the cleaning loss rate is: ; in, This represents the influence coefficient of the cleaning loss rate, which is related to the structural parameters of the fish scale screen and the fan.
[0046] In step 2, the measured data of the operating parameters and the corresponding measured data of the operating quality of the combine harvester under various combinations of operating parameters are obtained to form a dataset, specifically: Breakage rate and impurity content detection sensors (such as...) Figure 5 (As shown) It is installed on the outside of the combine harvester at the grain conveyor. The driving motor in the sampling device, based on the sampling time, and the industrial camera in the detection device acquire and process images. Impurities and breakage information are extracted based on different features such as pixel area. The breakage rate and impurity content calculation models are then called to output the average breakage rate and impurity content.
[0047] The entrainment loss rate sensor is installed on both sides below the threshing concave screen at the tail of the separation section of the threshing device (e.g., Figure 6 As shown), the cleaning loss rate sensor is installed at the tail of the cleaning device and below the straw crushing device (e.g. Figure 7 (As shown). The loss rate sensor amplifies, filters, and extracts features from the piezoelectric signal of material colliding with the sensor's sensitive plate, analyzing its amplitude, duration, and frequency to distinguish between grains and impurities. Subsequently, it combines the number of grains identified per unit time with parameters such as harvested area and yield for calculation.
[0048] Electric adjustment device for threshing gap (such as) Figure 8(As shown) It consists of an adjustable threshing screen, an electric push rod, a rotating shaft, connecting parts, and a connecting arm. When the electric push rod actuates, it drives the rotating shaft through the connecting rod, thereby driving the connecting arm to move and adjust the threshing screen. This adjustment structure is similar to a crank-rocker mechanism, converting the linear motion of the electric push rod into a change in the threshing gap through the connecting parts and the connecting arm. The adjustable range of the threshing gap is 10mm~30mm.
[0049] Fish scale screen opening electric adjustment device (such as...) Figure 9 (As shown) It consists of an electric push rod, a connecting rod, and an adjustable screen plate. The electric push rod is fixed to the fish scale screen frame, and the screen plate is connected to the end of the connecting rod and locked with a nut. This scheme converts the horizontal extension and retraction motion of the electric push rod into a change in the screen opening, with an adjustable screen opening range of 14mm to 26mm.
[0050] Amplitude-modulated PRBS was used as the input signal, and the input range of operating parameters was divided into equal intervals. A random number generator was used to sequentially select input parameter values within this range, collecting 70 discrete output data points. A Simulink module in MATLAB was used for data parsing, storage, and transmission. Operating speed, rolling speed, threshing gap, fan speed, and screen opening adjustment data were sent to the combine harvester line by line, while operating parameters and quality data were received from the combine harvester's CAN bus.
[0051] In step 3, a multi-strategy improved whale optimization algorithm is constructed, and the unknown parameters in the mechanistic model are identified using the dataset to determine the optimal model parameters, minimizing the error between the job quality predicted by the mechanistic model based on the job parameters and the measured job quality data. Specifically, the multi-strategy improved whale optimization algorithm integrates a Tent chaotic mapping initialization population strategy, a nonlinear adaptive convergence factor strategy, a nonlinear inertia weight strategy, a Lévy flight mechanism, a sparrow search algorithm development mechanism, and a Cauchy-Gaussian mutation strategy. The multi-strategy improved whale optimization algorithm process is as follows: Figure 4 As shown.
[0052] The improved Tent mapping method is calculated using the following formula: ; in, This represents a random number between the interval [0, 1].
[0053] The formula for calculating the nonlinear adaptive convergence factor is: ; in, Indicates the current iteration number; Indicates the maximum number of iterations. This represents the correlation coefficient between the iterative weights of adjustment and development. When... At that time, the overall trend of convergence factor change was relatively good.
[0054] The expression for calculating nonlinear weights is: ; The Levy flight mechanism calculation expression is: ; in, and Let be the direction vectors, respectively obeying and distributed, Generally take .
[0055] The location update formula for the SSA development mechanism is: ; in, Indicates the worst position globally. Indicates population size. Indicates the first The optimal position occupied by the discoverer in +1 iterations These are random numbers that follow a normal distribution. Represent a A matrix, where each element is randomly assigned the value 1 or -1, and . Represent a A matrix where each element is 1. When At this time, participants with poor fitness scores need to learn from other individuals to improve their fitness; when At that time, participants with higher fitness values will search around the current optimal position.
[0056] The formula for calculating the perturbation probability of the SSA development mechanism is: ; Cauchy-Gaussian mutation strategy calculation formula: ; in, It is a new solution after mutation. It is the position before the mutation; This represents the standard Cauchy distribution function. Indicates a Gaussian distribution; and Represents the adaptive mutation dynamic parameters. , .
[0057] Parameter identification involves using an improved whale optimization algorithm to identify unknown parameters of the threshing and cleaning operation quality mechanism model, and then employing reasonable evaluation indicators to assess the model's reliability. The principle of parameter identification can be expressed as follows: ; in, Indicates the test value. This represents the predicted value from the mechanistic model.
[0058] The formula for calculating the evaluation index of parameter identification is as follows: ; ; ; in, Indicates the number of data samples; This represents the mean of the experimental data. Indicates the first The mechanistic model predicts the value of each sample; RMSE The smaller the value, the smaller the error fluctuation between the identified model and the actual model. R 2 The larger the value, the closer the identified model is to the actual system, and the model output has the same trend as the experiment.
[0059] Based on the improved MIWOA (Minimum Ingress Whale Optimization) algorithm, the parameters of the mechanism models for breakage rate, impurity content, entrainment loss rate, and cleaning loss rate were identified. The iteration step count was set to 200 steps, and the parameter identification iteration curves are shown below. Figure 10 As shown in a, b, c, and d. The MIWOA algorithm was compared with the original WOA algorithm, and the parameter identification results were obtained after running 10 times. The results are shown in Table 1.
[0060] Table 1. Model parameter identification results
[0061] Based on the identified mechanistic model parameters, the output results of the mechanistic model parameter identification are as follows: Figure 11 As shown in a, b, c and d, the results of the mechanism model identification have a similar trend to the experimental values. The identified mechanism model can reflect the output response characteristics after the operation parameters change.
[0062] Based on the parameter identification results of the MIWOA algorithm, the calculation expression of the threshing and cleaning operation quality mechanism model is obtained: .
[0063] In step 4, validation data independent of the dataset is used to validate the prediction accuracy of the mechanistic model whose parameters have been identified. Specifically: The identified mechanism model was validated by re-evaluating the feed rate, drum speed, threshing gap, blower speed, and screen opening using a random number generator to obtain amplitude-modulated PRBS input signals different from those identified in the model. Output data was then collected at the same test site. The validation results of the threshing and cleaning mechanism model parameters, substituted into the identified mechanism model, are as follows: Figure 12 As shown in Figures a, b, c, and d. Model validation results show that the mechanism model for the quality of threshing and cleaning operations is valid. R 2 All are greater than 0.7. R All are greater than 0.85, while RMSE The values are relatively small, and the predicted values of the model are basically consistent with the trends of the experimental values, indicating high fitting accuracy. This shows that the established mechanistic model can well reflect the influence of operating parameters on the quality of threshing and cleaning operations, and has good predictive ability and generalization performance.
[0064] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0065] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0066] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1The steps of the function specified in one or more boxes.
[0067] Contents not described in detail in this specification are prior art known to those skilled in the art. It is hereby indicated that the above description is intended to help those skilled in the art understand this invention, but does not limit the scope of protection of this invention. Any equivalent substitutions, modifications, improvements, or simplifications of the above descriptions that do not depart from the essential content of this invention fall within the scope of protection of this invention.
Claims
1. A method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester, characterized in that, include: Step 1: Construct a threshing and cleaning operation quality mechanism model that characterizes the mapping relationship between operation parameters and operation quality; the operation parameters include at least the feeding speed, drum speed, threshing gap, blower speed and screen opening; the operation quality includes at least the grain breakage rate, entrainment loss rate, impurity content and cleaning loss rate; Step 2: Obtain measured data of operating parameters and corresponding measured data of operating quality of the combine harvester under various combinations of operating parameters to form a dataset; Step 3: Construct a multi-strategy improved whale optimization algorithm and use the dataset to identify unknown parameters in the mechanistic model to determine the optimal model parameters, minimizing the error between the job quality predicted by the mechanistic model based on the job parameters and the measured job quality data; wherein, the multi-strategy improved whale optimization algorithm integrates the Tent chaotic mapping initialization population strategy, the nonlinear adaptive convergence factor strategy, the nonlinear inertia weight strategy, the Lévy flight mechanism, the sparrow search algorithm development mechanism, and the Cauchy-Gaussian mutation strategy; Step 4: Validate the prediction accuracy of the mechanistic model with identified parameters using validation data independent of the dataset.
2. The method for identifying parameters of a combine harvester threshing and cleaning operation quality mechanism model according to claim 1, characterized in that, In step 1, a mechanism model for threshing and cleaning operation quality is constructed to characterize the mapping relationship between operation parameters and operation quality. Specifically: Based on the assumption of steady-state material flow mass conservation, the threshing and cleaning system of the combine harvester is regarded as a whole, and the mechanism modeling is carried out for the threshing process and the cleaning process respectively. In the modeling of the threshing process, a breakage rate mechanism model for calculating the grain breakage rate and an entrainment loss rate mechanism model for calculating the entrainment loss rate are constructed. In the modeling of the cleaning process, a material looseness coefficient model for the screen surface is constructed to characterize the looseness of the material. Based on the looseness coefficient model, a mechanism model for calculating the impurity rate and a mechanism model for calculating the cleaning loss rate are constructed respectively.
3. The method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester according to claim 2, characterized in that, Construct a breakage rate mechanism model, specifically as follows: A differential equation for grain crushing quality was established by deriving and relating the material layer volume fraction and the average number of collisions. Based on the relationship between threshing time and the axial velocity of the material flow, the differential equation is solved to obtain the calculation expression for the grain breakage rate. The calculation expression characterizes the grain breakage rate as a power function including the feeding speed, drum speed and threshing gap, and a comprehensive influence coefficient composed of material properties and structural parameters.
4. The method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester according to claim 3, characterized in that, Constructing a model for the entrainment loss rate mechanism, specifically: A differential equation for the mass of grain entrainment loss is established, taking the condition that the axial velocity of the material is greater than the radial velocity. The differential equation is solved based on the threshing time to obtain the calculation expression for the grain entrainment loss rate. The calculation expression characterizes the grain entrainment loss rate as a power function including the feeding speed, drum speed and threshing gap, as well as a comprehensive influence coefficient composed of material properties and structural parameters.
5. The method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester according to claim 4, characterized in that, The model for the bulk density coefficient of the material on the screen surface is constructed as follows: Construct calculation expressions for ideal material layer thickness, material layer thickness under vibration, and material layer thickness under airflow. Based on the material layer thickness under vibration and airflow, an expression for calculating the material layer thickness under the coupled action of vibration and airflow is constructed. The ratio of the material layer thickness under coupling effect to the ideal material layer thickness is defined as the material looseness coefficient of the screen surface, in order to correlate the influence of fan speed and screen opening on the looseness of the material.
6. The method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester according to claim 5, characterized in that, The impurity rate mechanism model and the cleaning loss rate mechanism model are constructed as follows: The calculation expression for the material looseness coefficient on the screen surface is substituted into the differential equations for impurity mass and grain loss mass established based on the concept of mass flow, respectively. By solving the differential equations and substituting variables based on the cleaning time, the calculation expressions for the impurity rate and the cleaning loss rate are obtained. The impurity rate calculation expression characterizes the impurity rate as a function of the feed rate, blower speed, screen opening, and the impurity rate influence coefficient. The cleaning loss rate calculation expression characterizes the cleaning loss rate as a function of the feed rate, blower speed, screen opening, and the cleaning loss rate influence coefficient.
7. The method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester according to claim 1, characterized in that, In step 2, the measured data of the operating parameters and the corresponding measured data of the operating quality of the combine harvester under various combinations of operating parameters are obtained, specifically as follows: Amplitude-modulated pseudo-random binary sequence signal is used as excitation signal to automatically adjust the operating parameters of the combine harvester over a wide range and at multiple sample points. The system collects real-time operating parameters and corresponding real-time operating quality data of the combine harvester under the excitation signal.
8. The method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester according to claim 7, characterized in that, The breakage rate and impurity rate in the real-time operation quality data are obtained through an image recognition method. The image recognition method includes: obtaining samples through a sampling device, acquiring sample images using an industrial camera, analyzing the pixel area features of particles in the image to distinguish between whole grains, broken grains and impurities, and calling a calculation model to output the breakage rate and impurity rate. The entrainment loss rate and cleaning loss rate in the real-time operation quality data are obtained through piezoelectric sensors. The piezoelectric sensors detect and analyze the amplitude, duration, and frequency characteristics of the piezoelectric signals generated by the material impacting the sensitive plate to distinguish between grains and impurities, and calculate the loss rate based on statistical data and operation parameters.
9. The method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester according to claim 1, characterized in that, In step 3, a multi-strategy improved whale optimization algorithm is constructed, specifically as follows: An improved Tent chaotic mapping method was used to initialize the whale population location to enhance the diversity and uniformity of the initial population. A nonlinear adaptive convergence factor strategy is adopted to balance the algorithm's global exploration and local exploitation capabilities; Nonlinear inertial weights and Lévy flight mechanism are introduced during the position update process to enhance the algorithm's ability to escape local optima. The development mechanism of the sparrow search algorithm is integrated, and the position of poor individuals is updated by a specific perturbation probability to improve the overall optimization ability of the population. A Cauchy-Gaussian mutation strategy is applied to the current optimal solution to improve the accuracy of the solution and prevent premature convergence of the algorithm.
10. The method for identifying parameters of a quality mechanism model for threshing and cleaning operations in a combine harvester according to claim 9, characterized in that, In step 3, the unknown parameters in the mechanistic model are identified using the dataset to determine the optimal model parameters. Specifically: The model parameters to be identified are used as optimization variables, and a fitness function is constructed based on the error between the job quality predicted by the mechanism model and the measured job quality data. The whale optimization algorithm is improved by using multiple strategies for iterative optimization. When the fitness function value is minimized or the termination condition is met, the current model parameters are output as the optimal model parameters.