Safety control method for field edge management system of high-horsepower tractor

Abstract

The invention provides a safety control method for a field management system of a high-horsepower tractor, relates to the technical field of electric control of tractors, realizes real-time monitoring and control through double CAN (controller area network) buses, ensures operation safety and efficiency, adopts a multifunctional handle unit, operates through an activation button and a safety button, monitors the state of the button in real time, and prevents mistaken touch. The action sequence is recorded and called, and the operation path and speed of the tractor are optimized through real-time feedback of sensor data in combination with a PID controller and a fuzzy logic control algorithm. When the system is started, the vehicle control unit carries out hardware self-inspection and environmental risk assessment, and safe operation is ensured. And on the basis of a fault prediction and multi-level early warning mechanism of machine learning, downtime is reduced, and maintenance efficiency is improved. According to the invention, the response speed and reliability of the tractor field edge management system and the operation safety are obviously improved, the accident risk caused by misoperation is reduced, and the damage to people and machines during abnormity is avoided.

Description

Technical Field

[0001] The invention relates to the technical field of tractor electronic control, and in particular to a safety control method for a headland management system of a high-horsepower tractor. Background Art

[0002] The implements mounted on high-horsepower tractors are very complicated, especially when turning at the end of the field, the driver needs to perform tedious operations, usually canceling cruise, lifting up, closing PTO, engaging reverse, activating the reversing multi-way valve, engaging forward, lifting down, activating PTO and activating cruise, etc. These tedious operations can easily make the driver tired, and improper operation can also damage the implements.

[0003] Chinese invention patent application CN106054683A discloses a tractor field management control system and control method thereof. Its core control mode is still memory-based, and the recording operation still requires manual operation by the user, which is inefficient, has a complex system structure, and is difficult to troubleshoot after a fault.

[0004] Chinese invention patent application CN106527228A discloses a one-touch field management and control method suitable for tractor farmland operations. Its human-computer interaction input is a hard-wired button. When the connection is abnormal or the button fails, the system cannot respond immediately to ensure safety.

[0005] Therefore, a safety control method for a high-horsepower tractor headland management system is proposed. Summary of the invention

[0006] The problem to be solved by the present invention is to provide a safety control method for a high-horsepower tractor headland management system, so that users can safely operate the headland management, on the one hand reducing the complexity of operation, and on the other hand ensuring driving safety to avoid damage to people and equipment in abnormal situations.

[0007] In order to achieve the above object, the present invention adopts the following technical solution: a safety control method for a high-horsepower tractor headland management system, comprising the following steps:

[0008] S1: When the system starts, the multi-function handle unit performs a self-test to check whether the button connection status and function are normal. If an abnormality is detected, the vehicle controller immediately issues a visual and sound alarm through the instrument display unit. The vehicle controller communicates with the multi-function handle unit through the CAN2 bus, operates through the activation button and safety button on the multi-function handle unit, and monitors the button status in real time;

[0009] S2: Press and hold the activation button and the safety button at the same time to start recording. The instrument display unit will display the corresponding icon to prompt the user that the current action has been recorded. Press and hold again to save the recorded action sequence. During the recording process, the vehicle controller collects and records the tractor's status data in real time through sensors.

[0010] S3: Short press the activation button and the safety button at the same time to call the optimized action sequence. The display screen on the instrument display unit shows the corresponding icon to ensure that the user can call and execute the optimized action sequence in order;

[0011] S4: collects the tractor's status data in real time through sensors, calculates the error between the current actual speed and the reference speed, generates a control signal and sends it to the actuator through the CAN1 bus to adjust the tractor's speed and direction;

[0012] S5: From start to finish, the system performs hardware self-check and operating environment risk assessment status checks and responds through real-time bus communication to ensure that the tractor operates in a safe state:

[0013] The CAN1 bus communication: The vehicle controller sends a heartbeat signal to each node on the CAN1 bus, receives the response signal in real time, monitors the status of sensors and actuators, and if any abnormality is detected, the vehicle controller immediately issues a visual and sound alarm through the instrument display unit to alert the user;

[0014] The CAN2 bus communication: The vehicle controller receives the response signal in real time and monitors the button status by sending a heartbeat signal to the multi-function handle unit on the CAN2 bus. If any abnormality is detected, the vehicle controller immediately issues a visual and sound alarm through the instrument display unit to remind the user;

[0015] S6: The user briefly presses the activation button and the safety button again, the vehicle controller stops the execution of the current action sequence, and the display screen on the instrument display unit displays the stop icon;

[0016] S7: Through historical data and real-time monitoring, potential failure points are predicted, and visual and audible alarms are issued through the instrument display unit to alert the operator.

[0017] Furthermore, in step S1, the activation button and the safety button need to be triggered at the same time to prevent safety hazards caused by accidental touches.

[0018] Furthermore, the safety control method supports real-time fault diagnosis and communication based on CAN bus.

[0019] Furthermore, the safety control method uses a vehicle control unit, covering differential lock, power output, lift control and electronically controlled hydraulic output.

[0020] Furthermore, the safety control method uses CAN1 to communicate with the engine control unit, the chassis control unit and the instrument display unit for real-time data transmission.

[0021] Furthermore, in step S2, the user long presses the activation button and the safety button again, the recording indicator light goes out, the buzzer stops sounding, the recorded action sequence is stored in the vehicle controller, and the action sequence is read and edited through the CAN2 bus to ensure that the recorded action sequence is stored correctly.

[0022] Furthermore, in step S3, the user performs a series of actions to be recorded. After each action is completed, the display screen on the instrument display unit displays a corresponding icon to prompt the user that the current action has been successfully recorded, ensuring that the user knows that each action has been recorded by the vehicle controller.

[0023] Furthermore, the step S4 specifically includes the following sub-steps:

[0024] Error calculation: calculate the deviation between the current actual driving speed and the target speed;

[0025] PID controller output: Use PID controller to calculate control signal, including: proportional part: multiply the proportional coefficient by the error to reduce the error, integral part: multiply the integral coefficient by the accumulated value of the error over time to eliminate the steady-state error, differential part: multiply the differential coefficient by the rate of change of the error to predict the future error change; PID controller output: proportional part + integral part + differential part, adjust the response of the vehicle controller so that the actual output is close to or equal to the desired reference value;

[0026] Fuzzy logic control: Map the error and error change rate to fuzzy sets and generate fuzzy control signals based on the fuzzy rule base;

[0027] Control signal fusion: The control signal generated by the PID controller and the control signal generated by the fuzzy logic are weighted and fused to generate the final control signal: 0.7·control signal generated by the PID controller + (0.3)·control signal generated by the fuzzy logic;

[0028] Control signal sending: The final control signal is sent to the actuator through the CAN1 bus to adjust the speed and direction of the tractor.

[0029] Furthermore, in step S5, hardware self-check and operating environment risk assessment are performed, specifically including:

[0030] Hardware self-check: If all sensors and actuators are in normal state and CAN bus communication is normal, H(t) = 1;

[0031] If any sensor or actuator is in an abnormal state, H(t) = 0. If the CAN bus communication is abnormal, the vehicle controller issues a visual and audible alarm through the backup communication channel or the local alarm device.

[0032] Work environment risk assessment: If all environmental factors are within the safe range, R(t) = 1;

[0033] If any environmental factor exceeds the safety range, R(t) = 0;

[0034] Calculate the safety state function S(t): Calculate the safety state function S(t)=min[H(t),R(t)] based on the smaller value of the hardware self-test result H(t) and the environmental risk assessment result R(t), ensuring that S(t)≥0.8, where S(t) is the normalized safety state value. If S(t)<0.8, the instrument display unit displays an error and stops operation;

[0035] The result of the safety status function S(t) is sent to the instrument display unit via the CAN1 bus.

[0036] Further, the step S7 includes:

[0037] Data collection: Collect various types of sensor data and send them to the vehicle controller (VCU) through the CAN1 bus. The sensor data is collected once per second.

[0038] Data preprocessing: remove rows containing missing values; standardize the data using a standard scaler to have a mean of 0 and a standard deviation of 1;

[0039] Feature extraction: Use the rolling window method to calculate the statistical features of each sensor data; use the recursive feature elimination method to select features to reduce the complexity of the model and the risk of overfitting;

[0040] Model training: Divide the dataset into training set and test set, with a ratio of 70% training set and 30% test set; Use random forest classifier to train the model, set the parameters to 100 trees and random seed to 42; Use test set data to evaluate the performance of the model, and calculate the accuracy, recall rate, and F1 score indicators;

[0041] Model deployment: Load the trained random forest model; standardize the real-time sensor data and select features, then input them into the model for fault prediction; output the prediction results and issue visual and audible alarms through the instrument display unit to alert the operator;

[0042] Multi-level fault warning: Set up a multi-level fault warning mechanism to issue different levels of alarms according to the severity of the fault, helping operators take timely measures;

[0043] Feedback and optimization: Operators can confirm the accuracy of prediction results and provide feedback data; the system records the results of each prediction and the actual fault situation for subsequent model optimization; uses online gradient descent to update model parameters in real time; regularly evaluates model performance, adjusts model parameters or retrains the model based on the evaluation results, and sends fault prediction and self-recovery processing results to the instrument display unit via the CAN1 bus.

[0044] The present invention has the following beneficial effects:

[0045] 1. In the present invention, by adopting a dual-bus (CAN1 and CAN2) communication mechanism, efficient control of the tractor headland management system is achieved. The CAN2 bus is responsible for real-time communication with the multi-function handle unit, ensuring that the status of the activation button and the safety button can be quickly recognized by the system, effectively preventing safety hazards caused by accidental touches. At the same time, the CAN1 bus is responsible for the transmission of key control signals, ensuring the real-time and accuracy of control commands. This dual-bus design improves the response speed and reliability of the system, significantly improves the safety of operation, and reduces the risk of accidents caused by operational errors.

[0046] 2. In the present invention, by combining fuzzy logic control, the fuzzy logic control algorithm can adjust the control signal in real time to ensure that the tractor operates according to the optimal path and speed. The predictive maintenance technology analyzes historical data and real-time monitoring data to predict potential fault points and trigger a multi-level fault warning mechanism, which not only improves the stability and reliability of the system, but also reduces maintenance costs and unexpected downtime. The integrated hardware self-check and operating environment risk assessment mechanism significantly improves the safety of the high-horsepower tractor headland management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flow chart of a safety control method for a high-horsepower tractor headland management system proposed by the present invention;

[0048] Figure 2 The present invention provides a system component connection diagram for a safety control method for a high-horsepower tractor headland management system. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] In one embodiment: Figure 1-Figure 2As shown, the present invention proposes a safety control method for a high-horsepower tractor headland management system, comprising the following steps:

[0051] S1: When the system starts, the multi-function handle unit performs a self-test to check whether the button connection status and function are normal. If an abnormality is detected, the vehicle controller immediately issues a visual and sound alarm through the instrument display unit. The vehicle controller communicates with the multi-function handle unit through the CAN2 bus, operates through the activation button and safety button on the multi-function handle unit, and monitors the button status in real time;

[0052] S2: Press and hold the activation button and the safety button at the same time to start recording. The instrument display unit will display the corresponding icon to prompt the user that the current action has been recorded. Press and hold again to save the recorded action sequence. During the recording process, the vehicle controller collects and records the tractor's status data in real time through sensors.

[0053] S3: Short press the activation button and the safety button at the same time to call the optimized action sequence. The display screen on the instrument display unit shows the corresponding icon to ensure that the user can call and execute the optimized action sequence in order;

[0054] S4: collects the tractor's status data in real time through sensors, calculates the error between the current actual speed and the reference speed, generates a control signal and sends it to the actuator through the CAN1 bus to adjust the tractor's speed and direction;

[0055] S5: From start to finish, the system performs hardware self-check and operating environment risk assessment status checks and responds through real-time bus communication to ensure that the tractor operates in a safe state:

[0056] CAN1 bus communication: The vehicle controller sends heartbeat signals to each node on the CAN1 bus, receives response signals in real time, monitors the status of sensors and actuators, and if any abnormality is detected, the vehicle controller immediately issues visual and audible alarms through the instrument display unit to alert the user;

[0057] CAN2 bus communication: The vehicle controller sends a heartbeat signal to the multi-function handle unit on the CAN2 bus, receives the response signal in real time, and monitors the button status. If any abnormality is detected, the vehicle controller immediately issues a visual and sound alarm through the instrument display unit to remind the user;

[0058] S6: The user briefly presses the activation button and the safety button again, the vehicle controller stops the execution of the current action sequence, and the display screen on the instrument display unit displays the stop icon;

[0059] S7: Through historical data and real-time monitoring, potential failure points are predicted, and visual and audible alarms are issued through the instrument display unit to alert the operator.

[0060] System architecture and components

[0061] 1. Multifunctional handle unit:

[0062] Activation button and safety button: The activation button and safety button are integrated. By pressing the activation button and the enabling button on the main joystick at the same time; the instrument display unit buzzer sounds, prompting the instrument display that the instrument display is ready to start recording the control sequence. Execute the first control action of the sequence to be recorded: the corresponding symbol is displayed on the instrument. Execute the next control action: the corresponding symbol is displayed on the instrument display. Note: A maximum of four symbols can be displayed on the instrument display at the same time. In the case of a sequence with more control actions, the left and right arrows indicate that there are other control actions in the current sequence.

[0063] Integrated fault diagnosis: The handle unit has a built-in self-check program that runs automatically each time the system is started to check the status of sensors and actuators. If an abnormality is found, an alarm signal is sent to the vehicle controller via the CAN2 bus.

[0064] 2. System Components

[0065] Engine Control Unit (ECU): responsible for monitoring engine speed, temperature, oil pressure and other parameters, and communicating with the vehicle controller through the CAN1 bus.

[0066] Chassis Control Unit (BCU): manages the basic operations of the vehicle such as driving, steering, braking, etc., and communicates with the vehicle controller through the CAN1 bus.

[0067] Instrument Display Unit (IDU): Provides visual information of operating status, including symbol display, digital display and buzzer reminder. Symbol display includes activation icon, fault icon, safety icon, etc.

[0068] Electronically controlled hydraulic output unit (EHO): controls the actions of the hydraulic system, such as lifting, lowering, and side shifting, and receives instructions from the vehicle controller through the CAN1 bus.

[0069] Handle Control Unit (HCU): Integrates activation and safety buttons for starting and stopping action sequences and communicates with the vehicle controller via the CAN1 bus.

[0070] Vehicle Controller Unit (VCU): As the core control unit, it coordinates the operation of each sub-unit to ensure the safety and efficient operation of the overall system. It communicates with each unit through the CAN1 bus to receive and send control instructions.

[0071] In large farmlands, we use high-powered tractors to perform tasks such as sowing, fertilizing and harvesting. In order to improve the efficiency and safety of operations, we use a high-powered tractor field management system safety control method. Workflow principle:

[0072] 1. Button application and safety check:

[0073] The operator operates through the activation button and safety button on the multifunctional handle unit. The vehicle controller communicates with the multifunctional handle unit through the CAN2 bus and monitors the button status in real time. If the button connection is abnormal or faulty, the vehicle controller responds immediately. If the button connection is abnormal or faulty, the vehicle controller issues visual and sound alarms through the instrument display unit to remind the operator, ensuring the safety of the operator's operation and preventing the danger caused by misoperation.

[0074] 2. Security verification logic:

[0075] When the system starts, the hardware self-check function and the operating environment risk assessment function begin to work, and the sensor status check: read the data of each sensor, such as the engine temperature sensor reading T, the hydraulic system pressure sensor reading P, and the position sensor reading L.

[0076] Engine temperature T range: 0℃, 100℃, if T>100℃, it is marked as abnormal.

[0077] Hydraulic system pressure P range: 0bar, 200bar, if P>200bar, it is marked as abnormal.

[0078] Position sensor reading L range: 0m, 1000m, if Mark as an exception.

[0079] Actuator status check: read the data of each actuator, such as motor current I and hydraulic pump pressure H.

[0080] Motor current I range: 0A, 100A, if I>100A, it is marked as abnormal.

[0081] Hydraulic pump pressure H range: 0bar, 200bar, if H>200bar, it is marked as abnormal.

[0082] Communication status check: Check the CAN bus communication status.

[0083] Packet loss rate L loss Range: 0%, 5%, if L loss >5%, marked as abnormal.

[0084] Calculation of hardware self-test result H(t): If all sensors and actuators are in normal state and CAN bus communication is normal, then H(t) = 1; if any sensor or actuator is in abnormal state, or CAN bus communication is abnormal, then H(t) = 0.

[0085] Working environment risk assessment: GPS positioning: Get the current geographical location of the tractor G and determine whether it is within the predetermined working area G zone .

[0086] Scheduled operation area G zone Range: G min ,G max ,if Marked as high risk.

[0087] Climate data: Get the current and future weather conditions through the weather API, such as wind speed W and rainfall R.

[0088] Wind speed W range: 0m / s, 20m / s. If W>20m / s, it is marked as high risk.

[0089] Rainfall R range: 0mm, 50mm. If R>50mm, it is marked as high risk.

[0090] Operation history: Analyze past operation data to identify potential risk points, such as areas with complex terrain. complex , the fault prone area F high If there is a history of failures at the current work site, it is marked as high risk.

[0091] Calculation of environmental risk assessment result R(t): If all environmental factors are within the safe range, R(t) = 1. If any environmental factor exceeds the safe range, R(t) = 0.

[0092] Safety state function S(t) calculation: S(t) = min[H(t), R(t)], take the smaller value of the hardware self-test result H(t) and the environmental risk assessment result R(t), that is, S(t) = min[H(t), R(t)]. Ensure that S(t) ≥ 0.8, where S(t) is the normalized safety state value. If S(t) < 0.8, the instrument display unit displays an error and stops operating. Ensure that the tractor operates in a safe state to prevent accidents caused by hardware failure or environmental risks. The range of the safety state function S(t) is [0, 1], where S(t) ≥ 0.8 indicates a safe state. The result of the safety state function S(t) is sent to the instrument display unit (IDU) via the CAN1 bus.

[0093] 3. Recording action sequences

[0094] The operator long presses the activation button and the safety button to start recording. The recording indicator light on the instrument display unit lights up and the buzzer sounds, prompting the operator that recording has started.

[0095] The operator performs a series of complex operations in sequence, such as starting the engine, adjusting the speed, steering, etc. The vehicle controller collects the tractor's status data in real time through sensors, including speed, direction, time, etc., and records these data in the vehicle controller.

[0096] After each action is completed, the display screen on the instrument display unit displays the corresponding icon to prompt the operator that the current action has been successfully recorded. Ensure that the operator knows that each action has been recorded by the vehicle controller to avoid omissions or incorrect records. The recorded action sequence can be called by the operator for subsequent automated operations. The action sequence includes specific actions such as steering, acceleration, deceleration, etc., and the parameters of each action such as speed, direction, time, etc. are recorded.

[0097] 4. End recording of action sequence

[0098] The operator presses and holds the activation button and the safety button again. The recording indicator goes out. The buzzer stops sounding. The recorded action sequence is stored in the tractor's vehicle control unit (VCU). Make sure that the recorded action sequence is correctly stored for subsequent recall and editing. The stored action sequence can be used when it is recalled. The action sequence is stored in the vehicle control unit and can be read and edited via the CAN2 bus.

[0099] 5. Calling action sequence

[0100] Calling actions: The operator short presses the activation button and the safety button to call the optimized action sequence. The instrument display unit displays the corresponding icons to ensure that the user can call and execute the optimized action sequence in sequence. Each short press executes an action, and the instrument flashes to display the next action icon. Ensure that the operator can call and execute the optimized action sequence in sequence, improving the accuracy and efficiency of the operation. The operator can call and execute the optimized action sequence through the instrument display unit.

[0101] 6. Real-time feedback and control

[0102] During the process of calling the action sequence, various sensors are used to collect the tractor's status data in real time, including speed, position, temperature, and pressure.

[0103] Error calculation: Calculate the current actual speed V(t) and the reference speed V ref (t) is the error e(t) between them.

[0104] PID controller output: Calculate the output u of the PID controller based on the error e(t) PID (t), where Proportional coefficient K p =0.5, integral coefficient K i =0.1, differential coefficient K d =0.05;

[0105] Fuzzy logic control: The error e(t) and the error change rate Mapped to fuzzy sets, generate fuzzy control signal u according to fuzzy rule base fuzzy (t).

[0106] The fuzzy sets include "negative big" (NB), "negative small" (NS), "zero" (Z), "positive small" (PS), and "positive big" (PB).

[0107] Rule base example: If the error e(t) is negative and the error change rate is is positively small, then the output u(t) is positively large. If the error e(t) is zero and the error change rate is is zero, the output u(t) is zero. Control signal fusion: The control signal u generated by the PID controller is PID (t) and the control signal u generated by fuzzy logic fuzzy (t) Perform weighted fusion to generate the final control signal u final (t), that is, u final (t) = α·u PID (t)+(1α)·u fuzzy (t), where α = 0.7. The final control signal u final (t) is sent to the actuator through the CAN1 bus. Through real-time feedback and adjustment, the tractor is ensured to always maintain the optimal state during operation, improving the accuracy and safety of the operation. The final control signal u final (t) is used to adjust the speed and direction of the tractor. The final control signal u final (t) Range: For speed adjustment, the control signal range is -10km / h, 10km / h. For direction adjustment, the control signal range is -10°, 10°.

[0108] 7. Call action stop

[0109] The operator briefly presses the activation button and the safety button again. The vehicle controller stops the execution of the current action sequence.

[0110] The display on the instrument display unit shows the stop icon and the buzzer stops sounding. This ensures that the operator can quickly stop the operation in an emergency to ensure safety. The operator can stop the execution of the current action sequence through the instrument display unit. The action sequence stops executing and the tractor returns to standby mode.

[0111] 8. Fault prediction and feedback

[0112] Collect multiple types of sensor data once per second using temperature sensors, pressure sensors, position sensors, velocity sensors, acceleration sensors, current sensors, and vibration sensors.

[0113] Read sensor data from the database to form a DataFrame; delete rows containing missing values ​​to ensure data integrity; use the StandardScaler to standardize the data so that its mean is 0 and its standard deviation is 1.

[0114] The statistical characteristics of each sensor data, including mean, standard deviation, maximum and minimum values, are calculated using the rolling window method;

[0115] The most relevant features are selected using the Recursive Feature Elimination (RFE) method, reducing the complexity of the model and the risk of overfitting.

[0116] The dataset is divided into a training set and a test set, with a ratio of 70% training set and 30% test set; the model is trained using a Random Forest Classifier, with the parameters set to 100 trees and a random seed of 42; the performance of the model is evaluated using the test set data, and indicators such as accuracy, recall, and F1 score are calculated.

[0117] Save the trained random forest model to a file; load the model from the file for use in real-time prediction; normalize the real-time sensor data and select features before inputting it into the model for fault prediction; output the prediction results to the vehicle controller, which issues visual and audible alarms through the instrument display unit to alert the operator.

[0118] Multi-level fault warning: Set up a multi-level fault warning mechanism to issue different levels of alarms according to the severity of the fault, helping operators take timely measures;

[0119] Level 1 warning: Condition: When the vehicle controller detects a minor fault, the engine temperature T exceeds 90°C but does not exceed 100°C

[0120] The hydraulic system pressure P exceeds 180 bar but does not exceed 200 bar.

[0121] The motor current I exceeds 90A but does not exceed 100A.

[0122] The hydraulic pump pressure H exceeds 180 bar but does not exceed 200 bar.

[0123] Output: A yellow warning light is displayed on the instrument display unit and a mild audible alarm sounds to alert the operator to potential problems.

[0124] Level 2 warning: Conditions: When the vehicle controller detects a medium fault, such as:

[0125] The engine temperature T exceeds 100°C but does not exceed 110°C.

[0126] The hydraulic system pressure P exceeds 200 bar but does not exceed 220 bar.

[0127] The motor current I exceeds 100A but does not exceed 110A.

[0128] The hydraulic pump pressure H exceeds 200 bar but does not exceed 220 bar.

[0129] Output: An orange warning light is displayed through the instrument display unit and a moderate audible alarm is sounded, advising the operator to check and handle the fault as soon as possible.

[0130] Level 3 warning: Condition: When the vehicle controller detects a serious fault, such as:

[0131] The engine temperature T exceeds 110℃

[0132] The hydraulic system pressure P exceeds 220 bar.

[0133] The motor current I exceeds 110A.

[0134] The hydraulic pump pressure H exceeds 220 bar.

[0135] Output: A red warning light is displayed on the instrument display unit and a high-intensity audible alarm is sounded, instructing the operator to immediately stop the operation and perform troubleshooting.

[0136] Feedback and optimization: Operators can confirm the accuracy of prediction results and provide feedback data; the system records the results of each prediction and the actual fault conditions for subsequent model optimization; online gradient descent is used to update model parameters in real time to gradually improve the prediction accuracy of the model; the performance of the model is regularly evaluated, and model parameters are adjusted or the model is retrained based on the evaluation results.

[0137] If a potential failure is predicted, the vehicle controller issues visual and audible alarms through the instrument display unit to alert the operator and take appropriate preventive measures.

[0138] Reduce operational interruptions and improve tractor reliability and safety.

[0139] The operator can take measures in advance according to the prompts on the instrument display unit to avoid operation interruptions caused by potential failures.

[0140] Fault prediction results include the type and location of potential fault points.

[0141] The tractor status data is collected in real time and sent to the vehicle control unit (VCU) via the CAN1 bus.

[0142] Vehicle Controller Unit (VCU): Receives sensor data, performs fault prediction, and sends the results to the Instrument Display Unit (IDU) via the CAN1 bus.

[0143] Instrument Display Unit (IDU): Displays fault prediction results and prompt information.

Claims

1. A safety control method for a high-horsepower tractor headland management system, characterized in that: The following steps are involved: S1: When the system starts, the multi-function handle unit performs a self-test to check whether the button connection status and function are normal. If an abnormality is detected, the vehicle controller immediately issues a visual and sound alarm through the instrument display unit. The vehicle controller communicates with the multi-function handle unit through the CAN2 bus, operates through the activation button and safety button on the multi-function handle unit, and monitors the button status in real time; S2: Press and hold the activation button and the safety button at the same time to start recording. The instrument display unit will display the corresponding icon to prompt the user that the current action has been recorded. Press and hold again to save the recorded action sequence. During the recording process, the vehicle controller collects and records the tractor's status data in real time through sensors. S3: Short press the activation button and the safety button at the same time to call the optimized action sequence. The display screen on the instrument display unit shows the corresponding icon to ensure that the user can call and execute the optimized action sequence in order; S4: collects the tractor's status data in real time through sensors, calculates the error between the current actual speed and the reference speed, generates a control signal and sends it to the actuator through the CAN1 bus to adjust the tractor's speed and direction; S5: From start to finish, the system performs hardware self-check and operating environment risk assessment status checks and responds through real-time bus communication to ensure that the tractor operates in a safe state: The CAN1 bus communication: The vehicle controller sends a heartbeat signal to each node on the CAN1 bus, receives the response signal in real time, monitors the status of sensors and actuators, and if any abnormality is detected, the vehicle controller immediately issues a visual and sound alarm through the instrument display unit to alert the user; The CAN2 bus communication: The vehicle controller receives the response signal in real time and monitors the button status by sending a heartbeat signal to the multi-function handle unit on the CAN2 bus. If any abnormality is detected, the vehicle controller immediately issues a visual and sound alarm through the instrument display unit to remind the user; S6: The user briefly presses the activation button and the safety button again, the vehicle controller stops the execution of the current action sequence, and the display screen on the instrument display unit displays the stop icon; S7: Through historical data and real-time monitoring, potential failure points are predicted, and visual and audible alarms are issued through the instrument display unit to alert the operator.

2. The safety control method for a high-horsepower tractor headland management system according to claim 1 is characterized in that: In step S1, the activation button and the safety button need to be triggered at the same time to prevent safety hazards caused by accidental touches.

3. The safety control method for a high-horsepower tractor headland management system according to claim 1 is characterized in that: The safety control method supports real-time fault diagnosis and communication based on CAN bus.

4. The safety control method for a high-horsepower tractor headland management system according to claim 1 is characterized in that: The safety control method uses a vehicle control unit, covering differential lock, power output, lifting control and electronically controlled hydraulic output.

5. The safety control method for a high-horsepower tractor headland management system according to claim 1 is characterized in that: The safety control method uses CAN1 to communicate with the engine control unit, the chassis control unit and the instrument display unit for real-time data transmission.

6. The safety control method for a high-horsepower tractor headland management system according to claim 1 is characterized in that: In step S2, the user long presses the activation button and the safety button again, the recording indicator light goes out, the buzzer stops sounding, and the recorded action sequence is stored in the vehicle controller. The action sequence is read and edited through the CAN2 bus to ensure that the recorded action sequence is correctly stored.

7. The safety control method for a high-horsepower tractor headland management system according to claim 1 is characterized in that: In step S3, the user performs a series of actions to be recorded. After each action is completed, the display screen on the instrument display unit displays a corresponding icon to prompt the user that the current action has been successfully recorded, ensuring that the user knows that each action has been recorded by the vehicle controller.

8. The safety control method for a high-horsepower tractor headland management system according to claim 1 is characterized in that: The step S4 specifically includes the following sub-steps: Error calculation: calculate the deviation between the current actual driving speed and the target speed; PID controller output: Use the PID controller to calculate the control signal, including: proportional part: multiply the proportional coefficient by the error to reduce the error, integral part: multiply the integral coefficient by the accumulated value of the error over time to eliminate the steady-state error, differential part: multiply the differential coefficient by the rate of change of the error to predict future error changes; PID controller output: proportional part + integral part + differential part, adjust the response of the vehicle controller so that the actual output is close to or equal to the desired reference value; Fuzzy logic control: Map the error and error change rate to fuzzy sets and generate fuzzy control signals based on the fuzzy rule base; Control signal fusion: The control signal generated by the PID controller and the control signal generated by the fuzzy logic are weighted and fused to generate the final control signal: 0.7·control signal generated by the PID controller + (0.3)·control signal generated by the fuzzy logic; Control signal sending: The final control signal is sent to the actuator through the CAN1 bus to adjust the speed and direction of the tractor.

9. The safety control method for a high-horsepower tractor headland management system according to claim 1, characterized in that: In step S5, hardware self-check and operating environment risk assessment are performed, specifically including: Hardware self-check: If all sensors and actuators are in normal state and CAN bus communication is normal, H(t) = 1; If any sensor or actuator is in an abnormal state, H(t) = 0. If the CAN bus communication is abnormal, the vehicle controller issues a visual and audible alarm through the backup communication channel or the local alarm device. Work environment risk assessment: If all environmental factors are within the safe range, R(t) = 1; If any environmental factor exceeds the safety range, R(t) = 0; Calculate the safety state function S(t): Calculate the safety state function S(t)=min[H(t),R(t)] based on the smaller value of the hardware self-test result H(t) and the environmental risk assessment result R(t), ensuring that S(t)≥0.8, where S(t) is the normalized safety state value. If S(t)<0.8, the instrument display unit displays an error and stops operation; The result of the safety status function S(t) is sent to the instrument display unit via the CAN1 bus.

10. The safety control method of a high-horsepower tractor headland management system according to claim 1, characterized in that: The step S7 comprises: Data collection: Collect various types of sensor data and send them to the vehicle controller (VCU) through the CAN1 bus. The sensor data is collected once per second. Data preprocessing: remove rows containing missing values; standardize the data using a standard scaler to have a mean of 0 and a standard deviation of 1; Feature extraction: Use the rolling window method to calculate the statistical features of each sensor data; use the recursive feature elimination method to select features to reduce the complexity of the model and the risk of overfitting; Model training: Divide the dataset into training set and test set, with a ratio of 70% training set and 30% test set; Use random forest classifier to train the model, set the parameters to 100 trees and random seed to 42; Use test set data to evaluate the performance of the model, and calculate the accuracy, recall rate, and F1 score indicators; Model deployment: Load the trained random forest model; standardize the real-time sensor data and select features, then input them into the model for fault prediction; output the prediction results and issue visual and audible alarms through the instrument display unit to alert the operator; Multi-level fault warning: Set up a multi-level fault warning mechanism to issue different levels of alarms according to the severity of the fault, helping operators take timely measures; Feedback and optimization: Operators can confirm the accuracy of prediction results and provide feedback data; the system records the results of each prediction and the actual fault situation for subsequent model optimization; uses online gradient descent to update model parameters in real time; regularly evaluates model performance, adjusts model parameters or retrains the model based on the evaluation results, and sends fault prediction and self-recovery processing results to the instrument display unit via the CAN1 bus.

Images