Torque control method and system for large-tonnage carrier

Abstract

The invention discloses a torque control method and system for a large-tonnage carrier, and the method comprises the steps: controlling a drive motor in real time, enabling the torque output torque of the drive motor to be close to a target torque, and obtaining the steady-state time of the target torque of the drive motor; if the steady-state time of the target torque is not less than a steady-state threshold value, changing a control parameter of a PID (Proportion Integration Differentiation) controller; and if the steady-state time of the target torque of the driving motor controlled by the PID controller is smaller than the steady-state threshold value after the control parameters are changed, the PID controller controls the driving motor to output the torque according to the control parameters. The torque response time of the double-shaft double-driving-force large-tonnage carrying vehicle is short, and meanwhile the torque steady state control capacity is improved.

Description

Technical Field

[0001] The present invention relates to a torque control method and system for a large-tonnage transport vehicle, which are applied to the torque control of a large-tonnage transport vehicle with dual driving forces. Background Art

[0002] The need for rapid torque control in dual-drive trucks is crucial due to the stringent requirements for dynamic response and synchronization accuracy in large-tonnage transport systems. Traditional single-axle drive solutions, limited by motor size and load capacity, struggle to meet the high-load demands of large-tonnage trucks. While dual-axle, dual-drive systems significantly improve load capacity and driving force through the symmetrical arrangement of dual motors and dual drive axles, they introduce more complex control challenges. Under dynamic conditions (such as heavy-load starting, sudden stops, or cornering), the dual motors must maintain precise synchronization of torque output to avoid mechanical stress concentration or differential overload caused by uneven torque distribution, which can compromise system stability and service life. Furthermore, while the compact H-shaped layout optimizes space utilization, it also exacerbates heat dissipation limitations and electromagnetic interference. Lagged torque response or overshoot can lead to increased energy loss and localized temperature rise, further threatening the reliability of critical components. Existing fixed-parameter PID controllers often fail to quickly converge to a steady-state state due to prolonged settling times when faced with sudden load changes or road disturbances, resulting in reduced transport efficiency and increased energy consumption. By real-time monitoring of torque buildup time and dynamically adjusting control parameters, the rigidity defects of traditional methods can be effectively overcome, ensuring the rapid response and coordinated operation of the dual-drive system under complex working conditions, thereby fully leveraging the redundancy advantages and expansion potential of the dual-motor architecture and meeting the dual requirements of large-tonnage transporters for high dynamic performance and robustness.

[0003] Prior art, such as a torque control method for a combined drive system of a mining dump truck, disclosed on the China Patent Network (Application No. 202410288151.5), suffers from technical deficiencies in dual-drive transport trucks. Dynamic response hysteresis: Torque control based on a fixed power distribution strategy (e.g., the equal torque distribution method) lacks a real-time parameter adjustment mechanism, preventing rapid convergence to a steady state in the event of sudden load changes or road disturbances. For example, the initial torque distribution in step 3 uses an equal distribution method, requiring the system to passively adjust the power share (see step 8) when a single motor overheats or fails. This adjustment delay can easily lead to mechanical stress concentration. Control parameters are rigid: Relying on a table lookup to determine the optimal operating temperature (step 8) and a preset external characteristic curve (step 1), this method is unable to adapt to the nonlinear dynamic characteristics under complex operating conditions. A two-dimensional table lookup is used to match the temperature range, but parameter iterative optimization is not performed in conjunction with real-time torque characteristics, resulting in reduced efficiency under extreme operating conditions. Algorithm scalability is limited: Existing methods do not integrate intelligent algorithms for parameter self-optimization. For example, steps 4-7 use a linear formula to calculate power distribution, which cannot address the torque coupling effects of unstructured roads. Summary of the Invention

[0004] The present invention aims to improve the anti-interference ability of the PID controller when responding to sudden load changes or road disturbances by adaptively adjusting the control parameters of the PID controller.

[0005] Another object of the present invention is to solve the problem that the adjustment time is too long and the machine cannot quickly converge to a steady state, resulting in reduced handling efficiency and increased energy consumption.

[0006] Another object of the present invention is to improve the scalability of the torque control method by adaptively adjusting the control parameters of the PID controller.

[0007] In order to achieve the above-mentioned objectives, the present invention provides a torque control method for a large-tonnage transport vehicle, and the method steps include: controlling the drive motor so that the output torque of the dual-drive transport vehicle is close to the target torque, and obtaining the target torque steady-state time; if the target torque steady-state time is not less than the steady-state threshold, updating the control parameters of the PID controller; if after the control parameters are updated, the target torque steady-state time of the drive motor controlled by the PID controller is less than the steady-state threshold, then the PID controller controls the output torque of the drive motor according to the control parameters.

[0008] Furthermore, the update process of the control parameters includes: obtaining torque timing-related characteristics, and inputting the torque timing-related characteristics into a machine learning model; the machine learning model outputs a first control parameter in response to the torque timing-related characteristics; optimizing the first control parameter, obtaining the second control parameter, and obtaining the updated control parameter.

[0009] Furthermore, the optimization process of the first control parameter includes: generating several intermediate control parameters based on the first control parameter; iterating several of the intermediate control parameters through a global optimization method until a set number of iterations is reached; obtaining the optimal intermediate control parameter among the several intermediate control parameters, and obtaining the second control parameter.

[0010] Furthermore, the optimization process of the first control parameter includes: generating several intermediate control parameters based on the first control parameter; iterating several of the intermediate control parameters through a global optimization method until a set number of iterations is reached; obtaining the optimal intermediate control parameter among the several intermediate control parameters, and obtaining the second control parameter.

[0011] Furthermore, the machine learning model includes: a neural network; the neural network outputs the first control parameter in response to the timing-related features.

[0012] Furthermore, the torque time series characteristics include: a fluctuation standard deviation, which is obtained by obtaining the sum of the square values of the differences between the torque corresponding to each moment and the target torque value, and dividing the sum by the quantity value corresponding to each moment minus the value one.

[0013] Furthermore, the torque time series characteristics include: wavelet transform characteristics, which are obtained by performing wavelet transform on the torque time series through a scaling function.

[0014] Furthermore, the torque timing-related characteristics include: a torque rise rate, obtaining the time from 10% of the target torque to 90% of the target torque, obtaining the difference of the target torques, and dividing the difference by the time.

[0015] In order to achieve the above-mentioned objectives, the present invention provides a torque control system for a large-tonnage transport vehicle, the system comprising: a control parameter updating module, which updates the control parameters of the PID controller in the control module in real time according to the target torque steady-state time; the control parameter updating module transmits the control parameters to the control module in real time through a data transmission channel, and the controller module updates the control parameters of the PID control according to the control parameters.

[0016] Furthermore, the control module includes: a PID controller, which controls at least one drive motor to output torque; and a control parameter receiver, which obtains the control parameters in the control parameter update module.

[0017] Furthermore, the system includes: an execution module, which has at least a drive motor and a sensor; the number of drive motors is at least 2, and the number of torque sensors is at least 4, and the sensor transmits the torque timing information or environmental collection information to the control parameter update module through a data transmission channel.

[0018] The beneficial effect of the present invention is that it significantly improves the torque response speed and synchronization accuracy of the dual-axle, dual-drive large-tonnage transporter. Its beneficial effects are reflected in: the system can monitor the torque output characteristics in real time and adaptively adjust the PID parameters based on the intelligent algorithm, effectively overcoming the response hysteresis problem of the traditional fixed parameter strategy under sudden load changes or complex road conditions; the torque timing characteristics are extracted through a neural network and the control parameters are generated in combination with the global optimization method to achieve high-precision coordination and temperature-efficiency coupling management of the dual-motor torque output, avoiding mechanical stress concentration or component overheating due to uneven power distribution; at the same time, the redundant sensor network and fault-tolerant control mechanism ensure power continuity in the event of a single-point failure. Combined with the compact hardware layout and dynamic differential compensation strategy, it takes into account both large-tonnage load requirements and high dynamic performance in a limited space, providing reliable technical support for the stable operation of heavy-duty transport equipment under harsh working conditions such as mines and ports. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The present invention is a flow chart of a method.

[0020] Figure 2 A torque timing diagram.

[0021] Figure 3 A neural network framework diagram.

[0022] Figure 4 The present invention relates to a framework diagram of the system. DETAILED DESCRIPTION

[0023] Example 1: This example discloses a specific implementation process of the method according to the present invention. Figure 1 .

[0024] The specific implementation of the method involved in this embodiment is described as follows.

[0025] The specific implementation steps of the method involved in this embodiment include: controlling the drive motor in real time so that the output torque of the drive motor is close to the target torque, and obtaining the target torque steady-state time of the drive motor.

[0026] In the above-mentioned torque control related steps, the driving motor is controlled according to the PID controller. The PID controller controls the output torque of the driving motor in the following way: the PID controller obtains the difference between the output torque value of the driving motor and the target torque value, and generates a control signal for the driving motor according to the difference.

[0027] Generates expressions for control signals Among them, K p is the proportional control parameter; K i is the integral control parameter; K d is the differential control parameter; e(t) is the difference between the torque output by the drive motor and the target torque at time t, and the expression is e(t) = T t -T target , T t is the torque output by the driving motor at time t, T target is the target torque.

[0028] In the above-mentioned torque control related steps, the process of obtaining the target torque steady-state time of the driving motor is as follows.

[0029] The control system obtains the torque output by the driving motor at each moment in the information transmitted by the sensor in real time, and obtains the target torque change moment and the initial steady-state moment of the torque.

[0030] The target torque change timing is specifically the time at which the target torque is changed. At each timing, the target torque value of the drive motor is updated according to changes in the surrounding environment.

[0031] The initial steady-state moment of torque is specifically: at this moment, the difference between the output torque of the drive motor corresponding to this moment and the target torque accounts for less than 0.05 of the target torque, and after this moment, the fluctuation value related to the target torque of the drive motor is less than the fluctuation threshold.

[0032] In order to further illustrate the initial steady-state moment of torque, an example is given below.

[0033] In one example of this embodiment, see Figure 2 ,exist Figure 2 In the equation, if time t1 is the initial steady-state moment of the torque, the relationship between the torque output by the drive motor corresponding to this moment and the target torque should satisfy the following relationship.

[0034] The specific relationship is: After satisfying this relationship, the following conditions must also be met.

[0035] The specific condition is: after this moment, the target torque fluctuation value of the driving motor is less than the fluctuation threshold value, and the specific relationship is: Where ε is the fluctuation threshold; S is the total number of moments. For example, if the time t1 is 2 ms, the timing step is 0.5 ms, and the end time is 10 ms, the total number of moments is 17.

[0036] After obtaining specific values of the target torque change time and the torque initial steady-state time, the target torque steady-state time is obtained according to the target torque change time and the torque initial steady-state time.

[0037] The target torque steady-state time is calculated as follows: the value corresponding to the initial steady-state moment of the torque minus the target torque change moment.

[0038] The specific implementation steps of the method involved in this embodiment include: after obtaining the target torque steady-state time of the drive motor, updating the control parameters of the PID control according to the size relationship between the target torque steady-state time and the steady-state threshold.

[0039] In the above torque control related steps, if the target torque steady-state time is less than the steady-state threshold, it means that the torque response time of the drive motor meets the requirements, and the control parameters of the PID controller are not updated. The control parameters of the PID controller remain unchanged, and the torque control of the drive motor is performed with the original control parameters.

[0040] If the target torque steady-state time is not less than the steady-state threshold, the control parameters of the PID controller are updated so that the target torque steady-state time of the drive motor is less than the steady-state threshold and the target torque steady-state time of the drive motor is optimized.

[0041] The process of updating the control parameters of the PID controller is as follows.

[0042] The torque timing data of the drive motor is obtained from the historical information of the sensor, and the torque timing related features are obtained from the torque timing data of the drive motor.

[0043] It should be noted that the starting time of the torque time series data is the time when the target torque was last updated.

[0044] In this embodiment, the torque time series related features include: the fluctuation standard deviation of the torque time series data, the wavelet characteristics of the torque time series data, the torque rising rate of the torque time series data, and the sum of square errors of the torque time series data.

[0045] The standard deviation of the fluctuation of the torque time series data is specifically: the square difference between the output torque of the drive motor corresponding to all moments of the torque time series data and the target torque, divided by the total number of moments minus the value of one.

[0046] The wavelet features of the torque time series data are specifically as follows: performing wavelet transform on the torque time series data, and extracting the wavelet transform features after the wavelet transform.

[0047] In a specific implementation process of this embodiment, in the process of extracting the wavelet features of the torque time series data, the wavelet scaling function used is the Morlet wavelet.

[0048] In a specific implementation process of this embodiment, in the process of extracting the wavelet features of the torque time series data, the wavelet scaling function used is the Haar wavelet.

[0049] In a specific implementation process of this embodiment, in the process of extracting the wavelet features of the torque time series data, the wavelet scaling function used is the Mexican hat wavelet.

[0050] In a specific implementation process of this embodiment, in the process of extracting the wavelet features of the torque time series data, the wavelet scaling function used is the Gaussian wavelet.

[0051] The torque rise rate of the torque timing data is specifically: subtract the numerical value of the moment corresponding to the actual output of the second proportion target torque of the drive motor from the numerical value of the moment corresponding to the actual output of the first proportion target torque of the drive motor to obtain the time difference, the difference between the actual output of the second proportion target torque of the drive motor and the actual output of the first proportion target torque of the drive motor to obtain the torque difference, and divide the torque difference by the time difference to obtain the torque rise rate.

[0052] In a specific implementation example of this embodiment, the moment corresponding to when the drive motor actually outputs 10% of the target torque and the moment corresponding to when the drive motor actually outputs 90% of the target torque are obtained, the numerical value of the moment corresponding to when the drive motor actually outputs 90% of the target torque is subtracted from the numerical value of the moment corresponding to when the drive motor actually outputs 10% of the target torque to obtain the moment difference, the difference between the actual output of 90% of the target torque by the drive motor and the actual output of 10% of the target torque by the drive motor is obtained to obtain the torque difference, and the torque difference is divided by the moment difference to obtain the torque rise rate.

[0053] In a specific implementation example of this embodiment, the moment corresponding to the actual output of 5% of the target torque by the drive motor and the moment corresponding to the actual output of 95% of the target torque by the drive motor are obtained, the numerical value of the moment corresponding to the actual output of 95% of the target torque by the drive motor is subtracted from the numerical value of the moment corresponding to the actual output of 5% of the target torque by the drive motor to obtain the moment difference, the difference between the actual output of 95% of the target torque by the drive motor and the actual output of 5% of the target torque by the drive motor is obtained to obtain the torque difference, and the torque difference is divided by the moment difference to obtain the torque rise rate.

[0054] In a specific implementation example of this embodiment, the moment corresponding to the actual output of 15% of the target torque by the drive motor and the moment corresponding to the actual output of 105% of the target torque by the drive motor are obtained, the numerical value of the moment corresponding to the actual output of 105% of the target torque by the drive motor is subtracted from the numerical value of the moment corresponding to the actual output of 15% of the target torque by the drive motor to obtain the moment difference, the difference between the actual output of 105% of the target torque by the drive motor and the actual output of 15% of the target torque by the drive motor is obtained to obtain the torque difference, and the torque difference is divided by the moment difference to obtain the torque rise rate.

[0055] The sum of square errors of the torque time series data is specifically: in the torque time series data, the sum of the square differences between the actual output torque of the drive motor and the target torque at each moment in the torque time series data is obtained.

[0056] Its expression E=∫e(t) 2 dt. Its discrete form is Where t0 is the initial time of the torque time series data, and t_end is the final time of the torque time series data. The meaning of e(t) has been explained in the previous section and will not be repeated here.

[0057] The steps of the specific implementation process of the method involved in this embodiment include: after obtaining the torque timing-related characteristics, inputting the torque timing-related characteristics into the machine learning model, and the machine learning model outputs a first control parameter in response to the torque timing-related characteristics.

[0058] In this embodiment, see Figure 3 The machine learning model is specifically a neural network. The acquired torque timing-related features are input into the neural network. The neural network outputs a first control parameter in response to the torque timing-related features. The first control parameter includes: a first proportional control parameter, a first integral control parameter and a first differential control parameter.

[0059] In a specific implementation process of this embodiment, the neural network is specifically a BP neural network, and the acquired torque training related features are input into the BP neural network, and the BP neural network responds to the torque timing related features and outputs the first control parameter.

[0060] After obtaining the first control parameter, several optimization vectors of the first control parameter are generated according to the first control parameter, and the optimization vectors of the several first control parameters are iterated through the global optimization algorithm. After reaching a certain number of iterations, the optimal optimization vector among the optimization vectors of the several first control parameters is obtained, and the first control parameter is optimized according to the optimal optimization vector to obtain the second control parameter.

[0061] The specific process of generating a plurality of optimized vectors of the first control parameters according to the first control parameters is as follows.

[0062] The optimization vector generation process involves randomly generating the value of the proportional control component corresponding to the first proportional control parameter, randomly generating the value of the integral control component corresponding to the first integral control parameter, and randomly generating the value of the differential control component corresponding to the first differential control parameter. The optimization vector is generated based on the values of the proportional control component, the integral control component, and the differential control component.

[0063] In the process of optimizing vector generation, the values of the proportional control component vector, the integral control component vector and the differential control component vector generated satisfy rK for the jth optimized vector. p，j 2 +rK i，j 2 +rK i，j 2 =1. Among them, rK p，jThe value of the proportional control component vector generated for the jth time; rK i，j The value of the integral control component vector generated for the jth time; rK d，j The value of the differential control component vector generated for the jth time.

[0064] For the jth optimization vector, the expression of the optimization vector is: Cv j =[rK p，j , rK i，j , rK d，j ].

[0065] The optimization vector generation process is performed several times until the number of generated optimization vectors reaches the set number.

[0066] The global optimization method is used to iterate over several optimization vectors until the number of iterations reaches the set number of iterations. The number of iterations is related to the number of optimization vectors, specifically the natural logarithm of the optimization vector multiplied by the adjustment factor C, rounded up.

[0067] Its expression is: in, For upward forensic functions, for example, Cv_sum is the number of generated optimization vectors.

[0068] In a specific implementation process of this embodiment, the value of the adjustment factor C is not less than 100 and not greater than 143. The adjustment factor C can take a value between 100 and 143.

[0069] In a specific implementation process of this embodiment, the value of the adjustment factor C is 101.

[0070] In a specific implementation process of this embodiment, the value of the adjustment factor C is 127.

[0071] In a specific implementation process of this embodiment, the value of the adjustment factor C is 142.

[0072] In this embodiment, the specific process of the global optimization method is as follows.

[0073] Step 1: Treat each different optimization vector as an individual. On this basis, obtain the global optimal individual among all individuals and obtain the historical optimal individual of the individuals.

[0074] The historical best individual of an individual is specifically: the individual with the highest score among the historical individuals when the individual is iterating.

[0075] The global optimal individual is specifically: the individual with the highest score among all current individuals.

[0076] Step 2: All individuals generate individual velocity vectors based on the historical optimal individual and the global optimal individual, and iterate the optimization vector corresponding to the individual based on the individual velocity vector.

[0077] The specific process of iterating the optimization vector corresponding to the individual is: update the velocity vector corresponding to all individuals. For the jth individual in the tth iteration process, the expression of the velocity iteration of the individual is: v t =w*v t-1 +I1*r1*(pbest j，t -Cv j，t )+I2*r2*(gbest t -Cv j，t ). Among them, I1 and I2 are individual learning factors and social learning factors respectively; r1 and r2 are random numbers in the range of (0,1); pbest j，t is the historical best individual of the jth individual at the tth iteration; gbest t is the global optimal individual at the tth iteration; Cv j，t is the jth individual in the tth iteration. After the speed iteration, the optimization vector corresponding to the best individual is iterated. For the jth individual, in the tth iteration process, the expression of the optimization vector iteration corresponding to the jth individual is: Cv j，t+1 =Cv j，t +v t .

[0078] Step 3: Update the historical optimal individual of all individuals and the global optimal individual of all individuals. If the number of iterations is less than the set number of iterations, execute step 2; if the number of iterations is not less than the set number of iterations, obtain the optimization vector corresponding to the individual with the highest score among all individuals.

[0079] In the specific process of the global optimization method, the process of obtaining each score is as follows.

[0080] In the historical data of the sensor, the output torque signal of the drive motor at each moment and the actual output torque are obtained, and the signal output ratio of the drive motor at each moment is generated according to the torque size corresponding to the output torque signal of the drive motor at each moment and the actual output torque size.

[0081] For the tth moment, the signal output ratio is specifically: the magnitude of the actual output torque of the driving motor at the tth moment divided by the magnitude of the torque corresponding to the output torque signal of the driving motor at the tth moment.

[0082] The optimization vector corresponding to each individual is added to the vector corresponding to the first control parameter. The vector corresponding to the first control parameter can be expressed as: K f =[K p，f , Ki，f , K d，f ], where K p，f , K i，f With K d，f They are respectively the first proportional control parameter, the first integral control parameter and the first differential control parameter.

[0083] Obtain the approximate torque time series data corresponding to each individual. For the jth individual, the approximate torque signal output by the driving motor at the tth moment is: Where eh(t) is the historical error signal, and its expression is eh(t)=Th t *AP t -T target Th t The torque value corresponding to the torque signal output by the driving motor at time t.

[0084] According to the torque time series data corresponding to each individual, the target torque steady-state time and the target torque related fluctuation value corresponding to the torque time series data are obtained. The process of obtaining the target torque steady-state time and the target torque related fluctuation value has been described in the above content and will not be repeated here.

[0085] The score of each individual is generated based on the target torque steady-state time and the target torque related fluctuation value of the torque time series data corresponding to each individual. The score of each individual is specifically: the steady-state threshold divided by the target torque steady-state time and the fluctuation threshold divided by the target torque related fluctuation value is raised to the power of the natural logarithm.

[0086] For the jth individual, the expression is: Among them, Tw j is the target torque steady-state time of the torque time series data corresponding to the jth individual; δ j is the target torque steady-state time of the torque time series data corresponding to the jth individual; ε is the steady-state threshold; is the fluctuation threshold.

[0087] Step 4: Optimize the first control parameter according to the obtained optimization vector to obtain the second control parameter.

[0088] In step 4, the second control parameter is obtained by adding the vector corresponding to the first control parameter to the obtained optimization vector. The expression is: K gbest =K f +Cv gbest , where Cv gbest is the optimization vector corresponding to the individual with the highest score among all individuals.

[0089] Through the method of this embodiment, in the application scenario of torque control of large-tonnage transport vehicles, while ensuring a fast response time of torque control and solving the problems of reduced transport efficiency and increased energy consumption, the stability of torque output during the torque control process of the drive motor is improved, and the method involved in this embodiment controls torque in a variety of environmental conditions.

[0090] Example 2: This example discloses the system according to the present invention. Figure 4 .

[0091] The system involved in this embodiment includes: a control parameter updating module, a control module and an execution module.

[0092] The control parameter updating module updates the parameters of the PID controller in the control module according to the size relationship between the target torque steady-state time and the steady-state threshold. After the control parameters of the PID controller are updated, the updated parameters of the PID controller are transmitted to the control module through the data transmission channel.

[0093] The process of the control parameter updating module for updating the control parameters of the PID controller is consistent with the PID control parameter updating process involved in Example 1, and will not be repeated here.

[0094] After receiving the PID controller parameters from the control parameter updating module through the control parameter receiver, the control module updates the control parameters of the PID controller in the control module.

[0095] The control module saves the torque signal output by the PID controller at each moment in real time through the data storage unit.

[0096] The control module controls the output torque of the driving motor in the execution module through a PID controller. There are at least two driving motors in the execution module and at least four torque sensors in the execution module.

[0097] The PID controller in the control module controls a drive motor in the execution module.

[0098] The torque sensor in the execution module transmits numerical information related to the actual output torque of the drive motor to the data storage part of the control module in real time.

[0099] This system implements intelligent torque control for dual-drive transport vehicles through a modular architecture. Its operational process relies on the efficient collaboration of a control parameter update module, a control module, and an execution module. As the system's intelligent decision-making core, the control parameter update module continuously monitors the actual output torque data fed back by the torque sensor in the execution module. It triggers a parameter optimization mechanism by calculating the deviation between the target torque steady-state time and a preset steady-state threshold in real time. When the target torque steady-state time exceeds the threshold, the module initiates the parameter update process. It first extracts historical torque time series data from the data storage unit and uses feature extraction techniques to analyze torque dynamic characteristics, including key indicators such as rise rate, overshoot, and steady-state fluctuation. These features are then input into a neural network model to generate preliminary control parameters. An optimization algorithm is then used to perform a multi-objective optimization search, ultimately generating the optimal parameter combination appropriate for the operating conditions. After parameter calculation, the control parameter update module transmits the updated parameters to the control module via a high-speed data transmission channel, ensuring real-time and reliable information transmission.

[0100] After receiving the new parameters, the control module immediately performs the parameter switching operation, pauses the current control cycle to load the new parameters and restarts the control loop. At the same time, the built-in fault-tolerant mechanism verifies the rationality of the parameters to prevent abnormal values from affecting the stability of the system. During the period when the parameters are effective, the control module adopts a closed-loop control strategy, receives the target torque command and sensor feedback value in real time, calculates the error signal through the improved PID algorithm, and outputs the control command to the execution module. The data storage unit fully records the multi-dimensional operating parameters of the control process, including timestamps, target and actual torque, environmental status and other information, providing data support for subsequent analysis and model iteration. During this process, the control module dynamically adjusts the control strategy to ensure that the torque output converges quickly to the target value, while taking into account the system energy consumption and component life.

[0101] The execution module, serving as the physical execution layer, employs a dual-motor redundant design and a multi-sensor layout to ensure the reliability and accuracy of power output. When the control module issues a torque command, the motor controller converts the digital signal into a drive current, driving the motor through power electronics to achieve precise torque output. A high-precision torque sensor collects the actual torque of the output shaft in real time, which is then fed back to the control module after signal conditioning and digital processing, forming a closed-loop control circuit. Under abnormal operating conditions, such as a single motor failure or sensor failure, the system automatically activates the redundancy mechanism, rapidly compensating torque output with healthy components to maintain the vehicle's basic traction capability. The hardware design of the execution module fully considers environmental adaptability, enabling stable operation under complex operating conditions. Furthermore, through the coordinated optimization of the mechanical structure and control strategy, the mechanical stress and thermal load on key components are reduced.

[0102] The technical advantage of the entire system lies in its balance between dynamic response and steady-state accuracy. The control parameter update module implements adaptive parameter adjustment through intelligent algorithms, overcoming the limitations of traditional fixed-parameter strategies under complex operating conditions. The control module's efficient data processing and fault-tolerant mechanisms ensure the accurate execution of control instructions. The execution module's redundant design and high-precision sensor network enhance the system's reliability and robustness at the hardware level. Working together, these three elements enable the system to maintain rapid convergence and stable tracking of torque output despite challenges such as sudden load changes, road disturbances, or ambient temperature fluctuations, thereby meeting the dual requirements of high dynamic performance and long-term reliability for large-tonnage transporters.

[0103] Example 3: Based on Example 1, this example discloses an application scenario in which the method of this example is applied to a dual-axle, dual-drive, large-tonnage transport vehicle.

[0104] In a dual-axle, dual-drive, large-tonnage transporter, the basic structure of the dual-motor drive system is that the core of the system is two motors, named the first drive motor and the second drive motor respectively.

[0105] First, a first speed sensor is installed on the first drive motor and connected to drive axle 1 via the first reduction gearbox, ensuring that the motor's speed can be monitored in real time. Simultaneously, a second speed sensor is installed on the second drive motor and connected to the second reduction gearbox and drive axle to monitor the motor's speed. Each motor is equipped with a corresponding motor controller. The first motor controller serves as the primary motor controller, responsible for the operation of the first drive motor, while the second motor controller serves as the secondary motor controller for the second drive motor. To ensure efficient system operation, the first and second motor controllers transmit signals via the CAN bus, ensuring real-time data sharing and coordination.

[0106] Specifically, during travel, the first motor controller sends a speed command to the first drive motor. Upon receiving the command, the first drive motor adjusts its speed according to the first motor controller's instructions to achieve the set vehicle speed. Simultaneously, the first drive motor's speed signal is fed back to the first motor controller via the first speed sensor for real-time monitoring.

[0107] Furthermore, the speed command is simultaneously transmitted to the second motor controller, ensuring the synchronization of the two motors. This design not only improves the vehicle's power performance but also enhances its stability and safety under high loads, ensuring the vehicle can efficiently and reliably complete various handling tasks.

[0108] When the transporter is working, the first drive motor and the second drive motor provide driving force at the same time. This design ensures that the transporter can still maintain stable power output even when the load is heavy.

[0109] The motor controller precisely synchronizes the two drive forces to effectively propel large-tonnage trucks. This dual-motor configuration significantly increases the truck's carrying capacity, enabling it to handle heavier loads without sacrificing speed or maneuverability. Under high loads, the motor controller optimizes energy distribution, ensuring balanced power output from both motors, minimizing energy consumption and extending battery life.

[0110] Furthermore, the dual-motor configuration offers a degree of fault redundancy. Even if one motor fails, the other can still provide basic driving power, thus ensuring the safety and reliability of the transporter. This design enables the transporter to operate efficiently in a variety of complex environments, not only improving work efficiency but also reducing downtime caused by failures, ensuring continuous and stable operation.

[0111] The dual-drive axle design, each equipped with an independent motor, provides the truck with greater load capacity. This configuration evenly distributes the load across each axle when carrying heavy loads, preventing deformation or damage to a single axle due to excessive weight. This design not only increases the truck's load capacity but also enhances its stability when operating on complex terrain, ensuring safe and reliable operation.

[0112] Furthermore, the dual-drive axle configuration enables the truck to provide smoother power output even under heavy loads, significantly improving operational efficiency. This dual-drive axle design maintains high maneuverability when handling heavy loads, reducing the risk of failures due to overload. Overall, the dual-drive axle arrangement provides high-tonnage trucks with a robust load-bearing capacity, ensuring safe and stable operation in a variety of operating conditions.

[0113] The specific application process of the torque control method involved in Example 1 in a dual-axle dual-drive large-tonnage transport vehicle is described as follows.

[0114] The coordinated operation of the control parameter update module, the control module, and the execution module achieves high-precision torque synchronization and dynamic optimization of the dual drive motors. At system startup, the control module initializes the PID parameters of the primary and secondary motor controllers and establishes a communication link between the two controllers via the CAN bus, ensuring real-time interaction between speed commands and feedback signals.

[0115] After the transporter enters the working state, the drive motor of the execution module outputs power according to the target torque command. At the same time, four high-precision torque sensors collect the output torque of the dual drive axle shafts in real time, convert the analog quantity into a digital signal through the signal conditioning circuit, and transmit it to the data storage part of the control module through the redundant data transmission channel.

[0116] The control module continuously compares the deviation between the target torque and the actual torque, calculates the torque steady-state time under the current working conditions, and triggers the control parameter update process when it detects that the steady-state time exceeds the preset threshold.

[0117] The control parameter update module receives the historical torque time series data from the data storage unit and uses the sliding window technology to extract key dynamic features, including torque rise rate, steady-state fluctuation range, overshoot and other indicators reflecting the system response characteristics.

[0118] During the feature extraction process, the system decomposes the torque signal through wavelet transform, obtains the energy distribution characteristics at different time scales, and constructs a multi-dimensional feature matrix based on time domain statistics.

[0119] The matrix is input into the pre-trained neural network model. The model analyzes the mapping relationship between torque dynamic characteristics and PID parameters based on the deep learning method and outputs a preliminary control parameter combination.

[0120] To further optimize parameter adaptability, the system uses a global optimization algorithm to iteratively optimize the initial parameters. By simulating torque response curves under different parameter combinations, the system identifies the optimal solution that balances response speed, stability, and energy efficiency. After parameter optimization, the control parameter update module encapsulates the new parameters into a data packet and synchronizes them to the primary and secondary motor controllers via a high-speed communication bus. This ensures parameter version consistency between the two controllers and avoids torque imbalances caused by parameter asynchrony.

[0121] After receiving the updated parameters, the control module first pauses the current control cycle, backs up the running PID parameters to the non-volatile memory, and then loads the new parameters and restarts the control cycle.

[0122] During the loading process, the module's built-in validation mechanism performs a logical review of the parameter value range. If an out-of-limit value or illegal format is detected, it automatically rolls back to the last valid parameter set to ensure system safety. After the parameters take effect, the main and sub-controllers use an improved differential-first PID algorithm to perform closed-loop control: the main controller receives the target torque command from the host computer and, combined with real-time data feedback from its own motor speed sensor, calculates the error signal and generates the PWM control variable. Simultaneously, the main controller sends synchronization commands to the sub-controller via the CAN bus. The sub-controller generates a matching drive signal based on the same algorithm, ensuring strict synchronization of the phase and amplitude of the dual motor output torque.

[0123] During this process, the data storage unit continuously records full-dimensional operating data including timestamp, ambient temperature, motor speed, actual torque and PID parameters, providing a data basis for offline analysis and model retraining.

[0124] The actuator module's dual drive motors feature a compact H-shaped layout, rigidly connected to the drive axle via a reduction gearbox, enabling high torque output within a limited space. When the control module issues a drive command, the motor controller converts the digital control variable into a three-phase current signal, which then drives the permanent magnet synchronous motor via the IGBT power module.

[0125] Four cross-platform torque sensors monitor output shaft torque in real time. Their signals undergo differential amplification and digital filtering to eliminate electromagnetic interference and mechanical vibration noise, generating highly reliable feedback data. During cornering, the differential automatically adjusts the left and right wheel speed difference based on the steering angle. The control system dynamically adjusts the torque distribution ratio between the two motors. The primary controller analyzes the steering sensor signals to calculate the difference in torque demand between the inner and outer wheels. The secondary controller then fine-tunes the output torque accordingly, ensuring smooth differential operation with minimal stress.

[0126] If the system detects that the torque output of one motor is abnormal due to overheating or failure, the redundant mechanism is immediately activated. The healthy motor increases the output torque in a very short time and adjusts the differential lock coefficient to maintain the basic traction performance of the vehicle.

[0127] The core advantage of the entire control process is reflected in the multi-level adaptive mechanism.

[0128] Under steady-state conditions, the system maintains high-precision tracking of torque output through closed-loop PID control; when encountering sudden load changes or road disturbances, the parameter update module quickly generates optimized parameters based on intelligent algorithms, significantly shortening the torque convergence time; for component aging or environmental temperature drift during long-term operation, the neural network model updates weight parameters through offline training to continuously improve the adaptability of the control strategy.

[0129] Furthermore, the dual-controller architecture and redundant sensor network design ensure that the system maintains basic functionality even in the event of a single point of failure, significantly improving operational reliability. Through the integrated application of these technical solutions, the dual-axle, dual-drive, high-tonnage transporter achieves rapid response, precise synchronization, and efficient energy management of power output under complex working conditions, providing reliable technical support for heavy-duty material handling operations.

[0130] The present invention has explained its purpose, technical solutions and beneficial effects in depth through specific embodiments, but these embodiments are only used as examples to demonstrate the application of the invention and do not constitute a limitation on the scope of protection of the present invention. We explicitly point out that any reasonable modification, equivalent replacement or technical improvement under the guidance of the spirit and principles of the present invention should be included in the scope of protection of the present invention. This means that as long as these changes do not deviate from the core idea and basic function of the invention, they should be protected by patent rights. The scope of protection of the present invention should be broad, including all direct and obvious variants and non-obvious innovations that technical experts can reasonably deduce based on the disclosure of the present invention. This broad protection is intended to promote further research and development based on the present invention, while ensuring that its innovation and practicality are fully legally protected.

Claims

1. A torque control method for a large-tonnage transport vehicle, characterized in that: The method steps include: controlling the drive motor in real time so that the output torque of the drive motor approaches the target torque, and obtaining the target torque steady-state time of the drive motor; If the target torque steady-state time is not less than the steady-state time threshold, the control parameters of the PID controller are updated; If the target torque steady-state time of the drive motor controlled by the PID controller is less than the steady-state time threshold after the control parameter is updated, the PID controller controls the drive motor to output torque according to the control parameter.

2. The torque control method for a large-tonnage transport vehicle according to claim 1, characterized in that: The updating process of the control parameters includes: Obtaining torque timing-related features, and inputting the torque timing-related features into a machine learning model; The machine learning model outputs a first control parameter in response to the torque timing-related characteristics; Optimize the first control parameter, obtain the second control parameter, and obtain the updated control parameter.

3. The torque control method for a large-tonnage transport vehicle according to claim 2, characterized in that: The process of optimizing the first control parameter includes: generating a plurality of optimized vectors of the first control parameters according to the first control parameters; Iterating a plurality of the optimization vectors by a global optimization method until a set number of iterations is reached; An optimal optimization vector is obtained from a plurality of optimization vectors, and the first control parameter is optimized according to the optimal optimization vector to obtain the second control parameter.

4. The torque control method for a large-tonnage transport vehicle according to claim 2 or 3, characterized in that: The machine learning model includes: a neural network; the neural network outputs the first control parameter in response to the torque timing-related characteristics.

5. The torque control method for a large-tonnage transport vehicle according to claim 2 or 3, characterized in that: The torque timing characteristics include: The standard deviation of fluctuation is obtained by obtaining the sum of the square values of the differences between the torque corresponding to each moment and the target torque value, and dividing the sum by the quantity value corresponding to each moment minus the value 1.

6. The torque control method for a large-tonnage transport vehicle according to claim 2 or 3, characterized in that: The torque timing characteristics include: Wavelet transform features are obtained by performing wavelet transform on the torque time series data through a scaling function.

7. The torque control method for a large-tonnage transport vehicle according to claim 2 or 3, characterized in that: The torque timing-related characteristics include: Torque increase rate, obtaining the time from the target torque of the first ratio to the target torque of the second ratio, obtaining the difference of the target torques, and dividing the difference by the time; The numerical value of the second ratio is greater than the numerical value of the first ratio.

8. A torque control system for a large-tonnage transport vehicle, implementing the torque control method according to any one of claims 1 to 7, characterized in that: The system comprises: A control parameter updating module updates the control parameters of the PID controller in the control module in real time according to the target torque steady-state time; the control parameter updating module transmits the control parameters to the control module in real time through a data transmission channel, and the controller module updates the control parameters of the PID control according to the control parameters.

9. The torque control system of a large-tonnage transport vehicle according to claim 8, characterized in that: The control module includes: A PID controller, wherein the PID controller controls the output torque of at least one drive motor; A control parameter receiver is used to obtain the control parameters in the control parameter updating module.

10. A torque control system for a large-tonnage transport vehicle according to claim 8 or 9, characterized in that: The system comprises: An execution module, the execution module at least comprising a drive motor and a torque sensor; The number of drive motors is at least 2, and the number of torque sensors is at least 4. The sensors transmit the timing information of the actual output torque of the drive motor to the control parameter update module through the data transmission channel.

Images