Intelligent adjusting system for operation speed of harvester

Abstract

The invention relates to the technical field of harvesters, in particular to an intelligent harvester operation speed adjusting system which comprises a data acquisition module, a crop lodging detection and compensation module, a pesticide residue detection and avoidance module, a control and execution module and a display and feedback module. The data acquisition module detects environmental parameters in real time through a plurality of sensors; the crop lodging detection and compensation module dynamically adjusts the operation speed and path of the harvester; the pesticide residue detection and avoidance module is used for detecting pesticide residue conditions on crops and avoiding areas with high pesticide residues; the control and execution module generates a corresponding operation speed adjusting instruction; and the display and feedback module is used for displaying the current operation parameters and the adjustment state. According to the invention, the intelligent level and the overall operation benefit of agricultural mechanization are improved, and the accuracy, safety and high efficiency of harvesting operation are ensured.

Description

Technical Field

[0001] The invention relates to the technical field of harvesters, and in particular to an intelligent regulating system for the operating speed of harvesters. Background Art

[0002] With the development of agricultural mechanization, harvesters are playing an increasingly important role in crop harvesting. However, in the actual harvesting process, the complex and changeable field environment and the diversity of crop conditions have brought many challenges to the harvesting operation, especially in areas of lodged crops and pesticide residues. Traditional harvesters are difficult to deal with effectively, which often leads to crop losses, reduced harvesting quality and environmental pollution. Therefore, there is an urgent need for an intelligent system that can adjust the operating speed and path of the harvester in real time to ensure the efficiency and safety of the harvesting operation.

[0003] In the existing technology, although there have been some research and applications on the optimization of harvester operations, most systems lack comprehensive consideration of complex environmental parameters, making it difficult to achieve precise and dynamic adjustments. For example, for the treatment of fallen crops, existing technologies mostly rely on fixed paths and speeds and cannot respond to changes in the degree of lodging in real time; for the detection and avoidance of pesticide residues, there is also a lack of efficient detection methods and path optimization algorithms, which makes it difficult to effectively avoid high pesticide residue areas during operations. In addition, the existing systems have deficiencies in data processing and feedback mechanisms, and are unable to monitor and adjust the operating status in real time, affecting the overall efficiency and quality of the harvesting operation.

[0004] The purpose of the present invention is to provide an intelligent adjustment system for the operating speed of a harvester, so as to improve the harvesting efficiency and operating quality, reduce crop losses and environmental pollution, and enhance the intelligence level and overall benefits of agricultural mechanization. Summary of the invention

[0005] Based on the above objectives, the present invention provides an intelligent adjustment system for the operating speed of a harvester.

[0006] The intelligent adjustment system for the operating speed of the harvester includes a data acquisition module, a crop lodging detection and compensation module, a pesticide residue detection and avoidance module, a control and execution module, and a display and feedback module, wherein; The data acquisition module detects environmental parameters in real time through multiple sensors, including crop density, terrain undulations, and soil moisture, and preprocesses the detected environmental parameters, including cleaning, standardization, time series analysis, and spatial interpolation; The crop lodging detection and compensation module detects the state data of the field crops and dynamically adjusts the operation speed and path of the harvester in combination with the environmental parameters, specifically including: Lodging detection: Lodging sensors and cameras are used to detect crop status data, including lodging conditions, height data, and posture information; Lodging analysis: Analyze the status data in combination with environmental parameters to identify the area, degree and direction of lodging crops, evaluate the impact of lodging crops on harvesting operations, and generate corresponding compensation strategies; Path optimization: Generate the initial operation path based on the results of lodging analysis, and dynamically adjust the harvester's operation path based on the lodging situation detected in real time; Speed ​​regulation: Based on the results of lodging analysis, the operating speed of the harvester is dynamically adjusted, and the harvesting effect and speed regulation effect are monitored in real time; The pesticide residue detection and avoidance module detects the pesticide residue on the crops and automatically adjusts the harvesting strategy to avoid areas with high pesticide residues based on the detection results; The control and execution module generates corresponding operation speed adjustment instructions and executes corresponding speed adjustment operations; The display and feedback module displays the current operating parameters and adjustment status, and feeds back the execution status and operating status to the operator.

[0007] Optionally, the data acquisition module includes: Data collection: Crop density is detected through crop density sensors, terrain undulations are detected through terrain sensors, and soil moisture is detected through humidity sensors; Data cleaning: Use statistical methods to remove noise and outliers, set threshold ranges, and treat environmental parameters that exceed the range as outliers and remove them; Data standardization: convert environmental parameters of different dimensions into unified dimensions; Time series analysis: Use the moving average method to smooth the environmental parameters and process the time variation trend of environmental parameters; Spatial interpolation: Use the inverse distance weighted (IDW) method to fill in the spatial data gaps within the sensor detection range.

[0008] Optionally, the lodging detection includes: Lodging sensor collection: installed on the front or side of the harvester, it uses ultrasonic technology to measure the height and lodging angle of crops in real time, and obtains the lodging status and posture information of crops through multi-point measurement; Camera collection: installed at the front of the harvester to obtain real-time images of crops and extract crop height data and posture information through image processing algorithms; Data fusion: The crop status data obtained by the lodging sensor and the camera are integrated to generate comprehensive lodging status, height data and posture information.

[0009] Optionally, the lodging analysis includes: Identification of lodging areas: Using the detected crop height data and lodging conditions, the crops are divided into different areas through K-means clustering to identify the areas of lodging crops; Lodging degree calculation: The lodging degree of each crop is calculated using the lodging index based on the lodging angle and height data; Lodging direction identification: Based on the lodging angle data, the direction vector method is used to identify the lodging direction of the crop; Impact assessment: Combined with environmental parameters (crop density, terrain relief and soil moisture), the impact of lodging crops on harvesting operations is assessed and an impact assessment report is generated, expressed as: ; Among them, I is the impact index, D is the crop density, G is the terrain undulation, S is the soil moisture, , , is the weight coefficient; Compensation strategy generation: Based on the results of the impact assessment, corresponding compensation strategies are generated, including adjusting the operating speed, optimizing the harvesting path, and adjusting the harvesting angle.

[0010] Optionally, the path optimization includes: Initial operation path generation: Based on the results of lodging analysis, the Dijkstra algorithm is used to generate the initial operation path, which is expressed as: ; in, is the shortest path distance from node v to the starting point, is the shortest path distance from node u to the starting point, is the edge weight between node u and node v, S is the set of processed nodes; Path cost calculation: Calculate the cost of a path, expressed as: ; ; in, is the shortest path distance to node v, is the shortest path distance to node u, is the edge weight between node u and node v, and are the coordinates of node v and node u respectively, is the weight coefficient of lodging degree, is the lodging degree between node u and node v; Dynamic path adjustment: Based on the real-time detected lodging conditions and environmental parameters, the initial path is optimized in real time using the Bellman-Ford algorithm, which is expressed as: ; in, is the shortest path distance to node v, is the shortest path distance to node u, is the edge weight between node u and node v, and V is the set of all nodes.

[0011] Optionally, the speed adjustment includes: Speed ​​adjustment: According to the results of lodging analysis, combined with the degree of lodging, crop density and soil moisture, the operating speed is dynamically calculated and expressed as: ; Where V is the adjusted operating speed, is the basic operation speed, L is the lodging degree, D is the crop density, S is the soil moisture, , , is the corresponding weight coefficient; Real-time monitoring: Use sensors to monitor harvesting results and speed adjustment effects in real time, including crop harvesting height, harvesting loss rate and machine load; Feedback adjustment: According to the results of real-time monitoring, the operation speed is dynamically adjusted, which is expressed as: ; in, For the adjusted operating speed, is the current operating speed, k is the adjustment coefficient, For the target harvesting effect, For the actual harvesting effect.

[0012] Optionally, the pesticide residue detection and avoidance module includes: Pesticide residue detection: Spectral sensors and cameras are used to detect pesticide residues on the surface of crops in real time, and the pesticide residues are integrated to generate comprehensive pesticide residue information; Residue analysis: Analyze the fused pesticide residue information through the convolutional neural network (CNN) model to identify the residue area and concentration; Strategy adjustment: Based on the results of residue analysis, generate an operation path that avoids areas with high pesticide residues and dynamically adjust the operation speed; Real-time monitoring and feedback: Monitor pesticide residues in real time through monitoring sensors, and use feedback to adjust the operation path and speed.

[0013] Optionally, the pesticide residue detection includes: Spectral sensor collection: Installed on the front or side of the harvester, the near-infrared spectral sensor is used to detect the pesticide residues on the surface of the crop in real time and obtain quantitative pesticide residue data, which is expressed as: ; in, is the reflectance spectrum, is the reflected light intensity, is the incident light intensity; Image acquisition: The surface image of the crop is collected by a camera installed on the front of the harvester, and the visual features of pesticide residues are identified using image processing algorithms to obtain qualitative pesticide residue data, which is expressed as: ; Among them, F is the feature data after image processing, I is the collected image data, is the image processing algorithm; Data fusion: The pesticide residue data obtained by the spectral sensor and the camera are fused to generate comprehensive pesticide residue information.

[0014] Optionally, the convolutional neural network (CNN) model includes: Convolutional layer: The convolutional layer is used to extract the local features of the comprehensive pesticide residue information, and multi-scale convolution kernels are added to enhance the detection ability of residue areas of different scales, which is expressed as: ; in, is the kth convolution kernel at position The output of , x is the input data, w is the convolution kernel weight, b is the bias, f is the ReLU activation function, M and N are the convolution kernel sizes; Batch Normalization: A batch normalization layer is added after each convolutional layer, expressed as: ; ; in, is the normalized data, and are the mean and standard deviation of the small batch data, and is a learnable parameter, To prevent a small constant with a denominator of zero, y is the output after batch normalization; Pooling layer: Use global average pooling to reduce the dimension of the feature map, expressed as: ; in, is the output of channel c, H and W are the height and width of the feature map respectively, is the position on channel c The input value of Fully connected layer: The pooled feature map is expanded and input into the fully connected layer, and features are integrated and classified through multiple fully connected layers, which can be expressed as: ; Among them, y is the output feature, W is the weight matrix, x is the input feature, b is the bias, and f is the ReLU activation function; Residue area and concentration identification: Use the Softmax function for multi-classification to identify the pesticide residue concentrations in different areas, expressed as: ; in, is the probability of category i, is the score of category i, and C is the total number of categories.

[0015] Optionally, the policy adjustment includes: Avoidance path generation: Based on the residue analysis results, the A* algorithm is used to generate the optimal operation path to avoid high pesticide residue areas, which is expressed as: ; ; ; in, is the estimated total cost of node n, is the actual cost from the starting point to node n, is the heuristic estimated cost from node n to the target node, is the actual cost of the parent node p, is the edge weight from the parent node p to the current node n, and are the coordinates of node n and target node respectively; Speed ​​adjustment: Based on the residue concentration and environmental parameters, the operating speed of the harvester is dynamically adjusted, expressed as: ; in, For the adjusted operating speed, is the initial operation speed, C is the pesticide residue concentration, D is the crop density, G is the terrain undulation, , , is the corresponding weight coefficient; Dynamic adjustment: Dynamically adjust the path and speed based on real-time monitoring data.

[0016] Beneficial effects of the present invention: The present invention ensures the accuracy and consistency of data through real-time detection and preprocessing of environmental parameters, including crop density, terrain undulation and soil moisture, and provides a reliable basis for subsequent intelligent adjustment. By comprehensively considering multiple factors such as crop density, terrain undulation, soil moisture, lodging and pesticide residues, an intelligent and dynamically adjusted harvesting operation speed adjustment system is provided, which significantly improves the intelligence level of agricultural mechanization and the overall operation efficiency, and ensures the accuracy, safety and efficiency of the harvesting operation.

[0017] The present invention detects crop status data through lodging sensors and cameras, dynamically adjusts the operating speed and path of the harvester, and ensures that crops in the lodging area can also be efficiently harvested. Through lodging detection, lodging analysis, path optimization and speed adjustment, the lodging situation, height data and posture information of the crops are comprehensively considered to provide accurate lodging crop identification and compensation strategies, effectively optimize the harvesting effect and speed adjustment effect, and ensure the efficiency and stability of the harvesting operation.

[0018] The present invention detects the pesticide residues on the surface of crops in real time through spectral sensors and cameras, and uses a convolutional neural network model to analyze the fused data to identify the residue area and concentration. Combined with the results of the residue analysis, the strategy adjustment module generates an optimal operation path that avoids the high pesticide residue area, and dynamically adjusts the operation speed to ensure the safety and efficiency of the operation. The real-time monitoring and feedback module monitors the pesticide residues in real time through monitoring sensors, and adjusts the operation path and speed through feedback, thereby improving the flexibility and adaptability of the system, ensuring the continuity of the operation, reducing crop losses, and improving the intelligence and efficiency of agricultural production. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 A schematic diagram of system function modules according to an embodiment of the present invention; Figure 2 Schematic diagram of a crop lodging detection and compensation module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0022] like Figure 1-Figure 2 As shown, the intelligent adjustment system for the operation speed of the harvester includes a data acquisition module, a crop lodging detection and compensation module, a pesticide residue detection and avoidance module, a control and execution module, and a display and feedback module, wherein; The data acquisition module detects environmental parameters in real time through multiple sensors, including crop density, terrain undulations, and soil moisture, and pre-processes the detected environmental parameters, including cleaning, standardization, time series analysis, and spatial interpolation; The crop lodging detection and compensation module detects the status data of field crops and dynamically adjusts the operating speed and path of the harvester based on environmental parameters to ensure that crops in the lodging area can also be harvested efficiently. Specifically, it includes: Lodging detection: Lodging sensors and cameras are used to detect crop status data, including lodging conditions, height data, and posture information; Lodging analysis: Analyze the status data in combination with environmental parameters to identify the area, degree and direction of lodging crops, evaluate the impact of lodging crops on harvesting operations, and generate corresponding compensation strategies; Path optimization: Generate the initial operation path based on the results of the lodging analysis, and dynamically adjust the harvester's operation path based on the real-time detected lodging situation to ensure that the harvester can efficiently avoid or enter the lodging area for harvesting; Speed ​​adjustment: Based on the results of lodging analysis, the operating speed of the harvester is dynamically adjusted to ensure that the harvesting is carried out at an appropriate speed in the lodging area to reduce crop losses. At the same time, the harvesting effect and speed adjustment effect are monitored in real time; The pesticide residue detection and avoidance module detects pesticide residues on crops and automatically adjusts harvesting strategies based on the detection results to avoid areas with high pesticide residues, thereby improving the safety and quality of harvesting operations. The control and execution module generates corresponding operation speed adjustment instructions and executes corresponding speed adjustment operations; The display and feedback module displays the current operating parameters and adjustment status, and feeds back the execution status and operating status to the operator to ensure the normal operation and timely adjustment of the system; Through the above contents, the harvesting efficiency and crop quality are effectively improved, while crop losses and environmental pollution are reduced, the safety and accuracy of harvesting operations are ensured, and the intelligence level of agricultural mechanization and operation results are significantly improved.

[0023] The data acquisition module includes: Data collection: Crop density is detected through crop density sensors, terrain undulations are detected through terrain sensors, and soil moisture is detected through humidity sensors; Data cleaning: Use statistical methods to remove noise and outliers, set a threshold range, and treat environmental parameters that exceed the range as outliers and remove them, expressed as: ; Where X is the original data set, is a single data point, is the mean of the data, is the standard deviation of the data; Data standardization: Convert environmental parameters of different dimensions into a unified dimension for subsequent analysis, expressed as: ; in, is the standardized data, is the original data, and are the minimum and maximum values ​​of the data set X respectively; Time series analysis: Use the moving average method to smooth the environmental parameters and process the time variation trend of the environmental parameters, which can be expressed as: ; in, is the moving average at time t, It's time The original data, n is the size of the time window; Spatial interpolation: Use the inverse distance weighted (IDW) method to fill in the spatial data gaps within the sensor detection range, expressed as: ; in, is the predicted value at the interpolation point P, It is a known point The value of Interpolation point P and known point The distance between them, k is the power exponent of the distance, usually 2, and N is the number of known points; Through the above steps, the accuracy, consistency and completeness of the data are ensured, thus providing a reliable foundation for subsequent intelligent adjustments, improving the overall accuracy and efficiency of the system, and effectively improving work quality and productivity.

[0024] Lodging detection includes: Lodging sensor collection: installed on the front or side of the harvester, it uses ultrasonic technology to measure the height and lodging angle of crops in real time, and obtains the lodging status and posture information of crops through multi-point measurement; The calculation formula for the lodging angle is: ; in, is the falling angle, and are the heights of the two measuring points, is the horizontal distance between two measuring points; Camera collection: installed at the front of the harvester to obtain real-time images of crops and extract crop height data and posture information through image processing algorithms; The calculation formula for height data is: ; Where h is the actual height, H is the height from the camera to the ground, p is the pixel height measured in the image, and f is the focal length of the camera; Data fusion: The crop status data obtained by the lodging sensor and the camera are fused to generate comprehensive lodging status, height data and posture information to improve the accuracy and reliability of the data, which is expressed as: ; in, is the fused state estimate, is the state estimate of the previous step, is the Kalman gain, is the measurement value, and H is the measurement matrix.

[0025] Lodging analysis includes: Lodging area identification: Using the detected crop height data and lodging conditions, the crops are divided into different areas through K-means clustering, and the areas of lodging crops are identified, which are expressed as: ; Among them, J is the objective function, k is the number of clusters, n is the number of data points, is the data point belonging to the i-th class, is the center of the i-th class; Calculation of lodging degree: The lodging degree of each crop is calculated using the lodging index based on the lodging angle and height data, expressed as: ; Among them, L is the lodging index, N is the number of detection points, is the lodging angle of the i-th detection point, is the height of the i-th detection point; Lodging direction identification: Based on the lodging angle data, the direction vector method is used to identify the lodging direction of the crop, expressed as: ; in, is the lodging direction vector, is the lodging angle of the i-th detection point; Impact assessment: Combined with environmental parameters (crop density, terrain relief and soil moisture), the impact of lodging crops on harvesting operations is assessed and an impact assessment report is generated, expressed as: ; Among them, I is the impact index, D is the crop density, G is the terrain undulation, S is the soil moisture, , , is the weight coefficient; Compensation strategy generation: Based on the results of the impact assessment, corresponding compensation strategies are generated, including adjusting the operating speed, optimizing the harvesting path, and adjusting the harvesting angle to reduce the impact of lodging on the harvesting operation; Through the above method, environmental parameters can be combined to accurately identify the area, degree and direction of lodging crops, evaluate their impact on harvesting operations, generate corresponding compensation strategies, and ensure the efficiency and stability of harvesting operations.

[0026] Path optimization includes: Initial operation path generation: Based on the results of the lodging analysis, the Dijkstra algorithm is used to generate the initial operation path to ensure that the harvester can avoid the area with severe lodging or enter the lodging area at the best angle, which is expressed as: ; in, is the shortest path distance from node v to the starting point, is the shortest path distance from node u to the starting point, is the edge weight between node u and node v, S is the set of processed nodes; Path cost calculation: Calculate the cost of a path, expressed as: ; ; in, is the shortest path distance to node v, is the shortest path distance to node u, is the edge weight between node u and node v, which usually takes into account factors such as distance and lodging degree. and are the coordinates of node v and node u respectively, is the weight coefficient of lodging degree, is the lodging degree between node u and node v; Dynamic path adjustment: Based on the real-time detected lodging conditions and environmental parameters, the Bellman-Ford algorithm is used to optimize the initial path in real time to ensure that the harvester can adapt to environmental changes during operation and maintain efficient operation, which is expressed as: ; in, is the shortest path distance to node v, is the shortest path distance to node u, is the edge weight between node u and node v, and V is the set of all nodes; Through the above steps, it is ensured that the harvester can operate efficiently in areas with severe lodging. By comprehensively considering environmental parameters such as the degree of lodging, crop density and soil moisture, and adjusting the operation path in real time, the harvesting efficiency and operation quality are improved, crop losses are reduced, the continuity and adaptability of the operation are ensured, and the intelligence level and overall benefits of agricultural mechanization are improved.

[0027] Speed ​​adjustment includes: Speed ​​adjustment: According to the results of lodging analysis, combined with the degree of lodging, crop density and soil moisture, the operating speed is dynamically calculated and expressed as: ; Where V is the adjusted operating speed, is the basic operation speed, L is the lodging degree, D is the crop density, S is the soil moisture, , , is the corresponding weight coefficient; Real-time monitoring: Use sensors to monitor harvesting results and speed adjustment effects in real time, including crop harvesting height, harvesting loss rate and machine load; Feedback adjustment: According to the results of real-time monitoring, the operation speed is dynamically adjusted to ensure the best harvesting effect, which is expressed as: ; in, For the adjusted operating speed, is the current operating speed, k is the adjustment coefficient, For the target harvesting effect, For actual harvesting effect; Through the above method, the operating speed of the harvester can be dynamically adjusted based on the results of lodging analysis combined with real-time monitoring data to ensure that the harvester can operate at the optimal speed under different lodging degrees and environmental conditions, thereby improving harvesting efficiency and crop quality.

[0028] The pesticide residue detection and avoidance module includes: Pesticide residue detection: Spectral sensors and cameras are used to detect pesticide residues on the surface of crops in real time, and the pesticide residues are integrated to generate comprehensive pesticide residue information; Residue analysis: Analyze the fused pesticide residue information through the convolutional neural network (CNN) model to identify the residue area and concentration; Strategy adjustment: Based on the results of residue analysis, an operation path is generated to avoid areas with high pesticide residues, and the operation speed is dynamically adjusted to ensure operation safety; Real-time monitoring and feedback: Monitor pesticide residues in real time through monitoring sensors, and provide feedback to adjust the operation path and speed to ensure the continuity and safety of the operation; Through the above content, dynamic adjustment of operation path and speed improves the flexibility and adaptability of the system, ensures the continuity of operations, reduces crop losses, and improves the intelligence and efficiency of agricultural production.

[0029] Pesticide residue testing includes: Spectral sensor collection: Installed on the front or side of the harvester, the near-infrared spectral sensor is used to detect the pesticide residues on the surface of the crop in real time and obtain quantitative pesticide residue data, which is expressed as: ; in, is the reflectance spectrum, is the reflected light intensity, is the incident light intensity, different wavelengths The corresponding reflectance spectra are used to identify pesticide residues; Image acquisition: The surface image of the crop is collected by a camera installed on the front of the harvester, and the visual features of pesticide residues are identified using image processing algorithms to obtain qualitative pesticide residue data, which is expressed as: ; Among them, F is the feature data after image processing, I is the collected image data, is the image processing algorithm; Data fusion: Fuse the pesticide residue data obtained by the spectral sensor and the camera to generate comprehensive pesticide residue information; Through the above method, the pesticide residues on the crop surface can be detected in real time, and the data from different sensors can be fused to generate comprehensive pesticide residue information, providing reliable data support for subsequent operation path adjustment and speed regulation.

[0030] Image processing algorithms Specifically include: Image preprocessing: The surface image of the collected crop is converted into a grayscale image and Gaussian blur is applied to reduce noise, which is expressed as: ; in, , , are the pixel values ​​of the red, green, and blue channels respectively, , , are the corresponding weights respectively; Edge detection: Use the Canny edge detection algorithm to extract the edges in the image, expressed as: ; in, It's pixels The gradient strength, and is the high and low threshold; Color detection: detect the color range and identify the color characteristics of pesticide residues, expressed as: ; in, , , Pixels The red, green, and blue channel values ​​of , , is the color range; Morphological analysis: Identify shape features and enhance image features through morphological operations such as dilation and erosion, expressed as: ; ; in, represents the expansion operation, represents the corrosion operation, K is the structural element; Feature extraction: Combine color detection and edge detection results to extract the characteristic area of ​​pesticide residues, expressed as: ; Among them, F is the feature data after image processing.

[0031] The Convolutional Neural Network (CNN) model consists of: Convolutional layer: The convolutional layer is used to extract the local features of the comprehensive pesticide residue information, and multi-scale convolution kernels are added to enhance the detection ability of residue areas of different scales, which is expressed as: ; in, is the kth convolution kernel at position The output of , x is the input data, w is the convolution kernel weight, b is the bias, f is the ReLU activation function, M and N are the convolution kernel sizes; Batch Normalization: A batch normalization layer is added after each convolutional layer to speed up training and improve the stability of the model, expressed as: ; ; in, is the normalized data, and are the mean and standard deviation of the small batch data, and is a learnable parameter, To prevent a small constant with a denominator of zero, y is the output after batch normalization; Pooling layer: Use global average pooling to reduce the dimension of the feature map while retaining more global information, expressed as: ; in, is the output of channel c, H and W are the height and width of the feature map respectively, is the position on channel c The input value of Fully connected layer: The pooled feature map is expanded and input into the fully connected layer, and features are integrated and classified through multiple fully connected layers, which can be expressed as: ; Among them, y is the output feature, W is the weight matrix, x is the input feature, b is the bias, and f is the ReLU activation function; Residue area and concentration identification: Use the Softmax function for multi-classification to identify the pesticide residue concentrations in different areas, expressed as: ; in, is the probability of category i, is the score of category i, and C is the total number of categories; Through the above content, the advantages of image and spectral data are combined to more accurately detect and analyze pesticide residues, improve the accuracy and reliability of detection, provide accurate pesticide residue information for harvesting operations, and ensure the safety and efficiency of operations.

[0032] Strategy adjustments include: Avoidance path generation: Based on the residue analysis results, the A* algorithm is used to generate the optimal operation path to avoid high pesticide residue areas, which is expressed as: ; ; ; in, is the estimated total cost of node n, is the actual cost from the starting point to node n, is the heuristic estimated cost from node n to the target node, is the actual cost of the parent node p, is the edge weight from the parent node p to the current node n, and are the coordinates of node n and target node respectively; Speed ​​adjustment: Based on the residue concentration and environmental parameters, the operating speed of the harvester is dynamically adjusted to ensure safe and efficient operation, expressed as: ; in, For the adjusted operating speed, is the initial operation speed, C is the pesticide residue concentration, D is the crop density, G is the terrain undulation, , , is the corresponding weight coefficient; Dynamic adjustment: Dynamically adjust the path and speed based on real-time monitoring data to ensure safety and efficiency during the operation; Through the above content, the shortest path to avoid areas with high pesticide residues can be effectively planned. At the same time, the operating speed can be dynamically adjusted according to the residue concentration and environmental parameters to ensure that the harvester operates efficiently in a safe area. Through real-time monitoring data and dynamic adjustment algorithms, the path and speed can adapt to changes in the working environment, thereby improving the safety and efficiency of the operation.

[0033] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0034] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. The intelligent adjustment system for the operating speed of the harvester is characterized by: It includes data acquisition module, crop lodging detection and compensation module, pesticide residue detection and avoidance module, control and execution module and display and feedback module, among which; The data acquisition module detects environmental parameters in real time through multiple sensors, including crop density, terrain undulations, and soil moisture, and preprocesses the detected environmental parameters, including cleaning, standardization, time series analysis, and spatial interpolation; The crop lodging detection and compensation module detects the state data of the field crops and dynamically adjusts the operation speed and path of the harvester in combination with the environmental parameters, specifically including: Lodging detection: Lodging sensors and cameras are used to detect crop status data, including lodging conditions, height data, and posture information; Lodging analysis: Analyze the status data in combination with environmental parameters to identify the area, degree and direction of lodging crops, evaluate the impact of lodging crops on harvesting operations, and generate corresponding compensation strategies; Path optimization: Generate the initial operation path based on the results of lodging analysis, and dynamically adjust the harvester's operation path based on the lodging situation detected in real time; Speed ​​regulation: Based on the results of lodging analysis, the operating speed of the harvester is dynamically adjusted, and the harvesting effect and speed regulation effect are monitored in real time; The pesticide residue detection and avoidance module detects the pesticide residue on the crops and automatically adjusts the harvesting strategy to avoid areas with high pesticide residues based on the detection results; The control and execution module generates corresponding operation speed adjustment instructions and executes corresponding speed adjustment operations; The display and feedback module displays the current operating parameters and adjustment status, and feeds back the execution status and operating status to the operator.

2. The intelligent adjustment system for harvester operation speed according to claim 1 is characterized in that: The data acquisition module comprises: Data collection: Crop density is detected through crop density sensors, terrain undulations are detected through terrain sensors, and soil moisture is detected through humidity sensors; Data cleaning: Use statistical methods to remove noise and outliers, set threshold ranges, and treat environmental parameters that exceed the range as outliers and remove them; Data standardization: convert environmental parameters of different dimensions into unified dimensions; Time series analysis: Use the moving average method to smooth the environmental parameters and process the time variation trend of environmental parameters; Spatial interpolation: Use the inverse distance weighted method to fill in the spatial data gaps within the sensor detection range.

3. The intelligent adjustment system for harvester operation speed according to claim 1 is characterized in that: The lodging detection comprises: Lodging sensor collection: installed on the front or side of the harvester, it uses ultrasonic technology to measure the height and lodging angle of crops in real time, and obtains the lodging status and posture information of crops through multi-point measurement; Camera collection: installed at the front of the harvester to obtain real-time images of crops and extract crop height data and posture information through image processing algorithms; Data fusion: The crop status data obtained by the lodging sensor and the camera are integrated to generate comprehensive lodging status, height data and posture information.

4. The intelligent adjustment system for harvester operation speed according to claim 3 is characterized in that: The lodging analysis includes: Identification of lodging areas: Using the detected crop height data and lodging conditions, the crops are divided into different areas through K-means clustering to identify the areas of lodging crops; Lodging degree calculation: The lodging degree of each crop is calculated using the lodging index based on the lodging angle and height data; Lodging direction identification: Based on the lodging angle data, the direction vector method is used to identify the lodging direction of the crop; Impact assessment: Combined with environmental parameters, the impact of lodging crops on harvesting operations is assessed and an impact assessment report is generated, expressed as: ； Among them, I is the impact index, D is the crop density, G is the terrain undulation, S is the soil moisture, , , is the weight coefficient; Compensation strategy generation: Based on the results of the impact assessment, corresponding compensation strategies are generated, including adjusting the operating speed, optimizing the harvesting path, and adjusting the harvesting angle.

5. The intelligent adjustment system for harvester operation speed according to claim 4 is characterized in that: The path optimization includes: Initial operation path generation: Based on the results of lodging analysis, the Dijkstra algorithm is used to generate the initial operation path, which is expressed as: ； in, is the shortest path distance from node v to the starting point, is the shortest path distance from node u to the starting point, is the edge weight between node u and node v, S is the set of processed nodes; Path cost calculation: Calculate the cost of a path, expressed as: ; ； in, is the shortest path distance to node v, is the shortest path distance to node u, is the edge weight between node u and node v, and are the coordinates of node v and node u respectively, is the weight coefficient of lodging degree, is the lodging degree between node u and node v; Dynamic path adjustment: Based on the real-time detected lodging conditions and environmental parameters, the initial path is optimized in real time using the Bellman-Ford algorithm, which is expressed as: ； in, is the shortest path distance to node v, is the shortest path distance to node u, is the edge weight between node u and node v, and V is the set of all nodes.

6. The intelligent adjustment system for harvester operation speed according to claim 5 is characterized in that: The speed regulation includes: Speed ​​adjustment: According to the results of lodging analysis, combined with the degree of lodging, crop density and soil moisture, the operating speed is dynamically calculated and expressed as: ； Where V is the adjusted operating speed, is the basic operation speed, L is the lodging degree, D is the crop density, S is the soil moisture, , , is the corresponding weight coefficient; Real-time monitoring: Use sensors to monitor harvesting results and speed adjustment effects in real time, including crop harvesting height, harvesting loss rate and machine load; Feedback adjustment: According to the results of real-time monitoring, the operation speed is dynamically adjusted, which is expressed as: ； in, For the adjusted operating speed, is the current operating speed, k is the adjustment coefficient, For the target harvesting effect, For the actual harvesting effect.

7. The intelligent adjustment system for harvester operation speed according to claim 1, characterized in that: The pesticide residue detection and avoidance module includes: Pesticide residue detection: Spectral sensors and cameras are used to detect pesticide residues on the surface of crops in real time, and the pesticide residues are integrated to generate comprehensive pesticide residue information; Residue analysis: Analyze the fused pesticide residue information through the convolutional neural network model to identify the residue area and concentration; Strategy adjustment: Based on the results of residue analysis, generate an operation path that avoids areas with high pesticide residues and dynamically adjust the operation speed; Real-time monitoring and feedback: Monitor pesticide residues in real time through monitoring sensors, and use feedback to adjust the operation path and speed.

8. The intelligent adjustment system for harvester operation speed according to claim 7, characterized in that: The pesticide residue detection includes: Spectral sensor collection: Installed on the front or side of the harvester, the near-infrared spectral sensor is used to detect the pesticide residues on the surface of the crop in real time and obtain quantitative pesticide residue data, which is expressed as: ； in, is the reflectance spectrum, is the reflected light intensity, is the incident light intensity; Image acquisition: The surface image of the crop is collected by a camera installed on the front of the harvester, and the visual features of pesticide residues are identified using image processing algorithms to obtain qualitative pesticide residue data, which is expressed as: ； Among them, F is the feature data after image processing, I is the collected image data, is the image processing algorithm; Data fusion: The pesticide residue data obtained by the spectral sensor and the camera are fused to generate comprehensive pesticide residue information.

9. The intelligent adjustment system for harvester operation speed according to claim 8, characterized in that: The convolutional neural network model includes: Convolutional layer: The convolutional layer is used to extract the local features of the comprehensive pesticide residue information, and multi-scale convolution kernels are added to enhance the detection ability of residue areas of different scales, which is expressed as: ； in, is the kth convolution kernel at position The output of , x is the input data, w is the convolution kernel weight, b is the bias, f is the ReLU activation function, M and N are the convolution kernel sizes; Batch Normalization: A batch normalization layer is added after each convolutional layer, expressed as: ; ； in, is the normalized data, and are the mean and standard deviation of the small batch data, and is a learnable parameter, To prevent a small constant with a denominator of zero, y is the output after batch normalization; Pooling layer: Use global average pooling to reduce the dimension of the feature map, expressed as: ； in, is the output of channel c, H and W are the height and width of the feature map respectively, is the position on channel c The input value of Fully connected layer: The pooled feature map is expanded and input into the fully connected layer, and features are integrated and classified through multiple fully connected layers, which can be expressed as: ； Among them, y is the output feature, W is the weight matrix, x is the input feature, b is the bias, and f is the ReLU activation function; Residue area and concentration identification: Use the Softmax function for multi-classification to identify the pesticide residue concentration in different areas, expressed as: ； in, is the probability of category i, is the score of category i, and C is the total number of categories.

10. The intelligent adjustment system for harvester operation speed according to claim 9, characterized in that: The strategy adjustment includes: Avoidance path generation: Based on the residue analysis results, the A* algorithm is used to generate the optimal operation path to avoid high pesticide residue areas, expressed as: ; ; ； in, is the estimated total cost of node n, is the actual cost from the starting point to node n, is the heuristic estimated cost from node n to the target node, is the actual cost of the parent node p, is the edge weight from the parent node p to the current node n, and are the coordinates of node n and target node respectively; Speed ​​adjustment: Based on the residue concentration and environmental parameters, the operating speed of the harvester is dynamically adjusted, expressed as: ； in, For the adjusted operating speed, is the initial operation speed, C is the pesticide residue concentration, D is the crop density, G is the terrain undulation, , , is the corresponding weight coefficient; Dynamic adjustment: Dynamically adjust the path and speed based on real-time monitoring data.

Images