A double-motor cooperative control system and method of a double-driver type trackless rubber-tyred vehicle for mines

Abstract

The application relates to the technical field of vehicle motor control, and discloses a double-motor cooperative control system and method for a mine-used double-driver trackless rubber-tyred vehicle, which comprises the following steps: collecting running state data, driver operation data and roadway environment data of the mine-used double-driver trackless rubber-tyred vehicle in real time, analyzing the real-time state of the vehicle, and constructing a first state vector; configuring a double-motor cooperative control mechanism, analyzing the state constraint constructed by the first state vector, respectively calculating target torques of a first driving motor and a second driving motor, and obtaining a first torque instruction; respectively controlling the working processes of the first driving motor and the second driving motor, and collecting state data of the motors in real time; respectively sending the state data of the first driving motor and the second driving motor to a controller of a motor on the opposite side, comparing the rotating speeds of the two motors, calculating corresponding compensation torques, and optimizing the control process of the motors; and the application can improve the driving stability and controllability of the vehicle under complex working conditions, and guarantee the safety of underground personnel and equipment.

Description

Technical Field

[0001] This application relates to the field of vehicle motor control technology, and more specifically to a dual-motor cooperative control system and method for a dual-driver trackless rubber-tired vehicle used in mining. Background Technology

[0002] Traditional tunneling operations primarily rely on manual labor and conveyor belts for transportation, resulting in significant waste of manpower and resources. This also hinders the progress of tunneling and poses safety hazards, frequently leading to material detachment and loss, as well as material entanglement in the conveyor belt, causing damage to the equipment. Trackless rubber-tired mining vehicles, as core equipment for auxiliary transportation in coal mines, directly impact mine production efficiency and operational safety. Their dual-cab design enhances maneuverability and flexibility in narrow and complex tunnels, making them particularly suitable for transporting materials and personnel over long distances and through multi-branched tunnels.

[0003] The existing technology has the following problems: Using a single engine with a mechanical gearbox, drive axle, and mechanical differential for power transmission and distribution makes it difficult to achieve torque distribution between the left and right wheels; the mechanical transmission chain is long, resulting in significant energy loss and affecting engine efficiency; the control methods for single-motor or dual-motor drives are simple, employing open-loop speed synchronization or simple torque averaging strategies, which do not fully consider the actual conditions of the vehicle and tunnel environment. When the road surface adhesion coefficients on both sides are different or there are slight differences in motor characteristics, speed differences can easily occur, leading to internal friction, deviation, or instability; the control strategy is fixed and cannot adapt to dynamically changing environmental factors such as tunnel slope, curvature, and road surface slippage, resulting in low control efficiency and insufficient precision. To solve at least one of the above problems, this application proposes a dual-motor cooperative control system and method for a dual-driver trackless rubber-tired vehicle used in mining. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this application is to provide a dual-motor coordinated control system and method for a dual-driver trackless rubber-tired vehicle used in mining, which can effectively solve the problems in the background technology. The specific technical solution of this application is as follows:

[0005] A dual-motor coordinated control method for a dual-driver trackless rubber-tired vehicle used in mining includes:

[0006] Real-time data collection of operating status, driver operation data, and roadway environment data of dual-driver trackless rubber-tired vehicles for mining; analysis of real-time vehicle status; and construction of the first state vector.

[0007] A dual-motor cooperative control mechanism is configured, state constraints are constructed by analyzing the first state vector, and the longitudinal and lateral forces of the vehicle are coupled and analyzed. The target torques of the first drive motor and the second drive motor are calculated respectively to obtain the first torque command. The target torque includes the target drive torque and the target braking torque.

[0008] In response to the first torque command, the operation of the first drive motor and the second drive motor are controlled respectively, and the status data of the motors are collected in real time to obtain the status data of the two motors.

[0009] Based on the dual-motor status data, the status data of the first drive motor and the second drive motor are sent to the motor controller on the other side, the speeds of the two motors are compared and the corresponding compensation torque is calculated, and the motor control process is optimized according to the compensation torque in order to perform coordinated control of the dual motors.

[0010] Specifically, the real-time acquisition of operating status data, driver operation data, and roadway environment data of the dual-driver trackless rubber-tired mining vehicle, analysis of the vehicle's real-time status, and construction of a first state vector include:

[0011] Real-time acquisition of operating status data, driver operation data, and roadway environment data of dual-driver trackless rubber-tired vehicles for mining, resulting in multi-source vehicle data;

[0012] Multi-source vehicle data is synchronized in time and aligned in space. The geometric parameters of the roadway, vehicle pose parameters and environmental obstacle parameters are analyzed to construct the first state vector.

[0013] Specifically, the dual-motor cooperative control mechanism includes:

[0014] The first state vector is analyzed to construct state constraints. The longitudinal and lateral dynamics of the vehicle are coupled and analyzed to calculate the longitudinal demand force and yaw moment of the vehicle. The dynamic feasible region of the dual motor joint operation is calculated based on the state constraints.

[0015] Target points are constructed based on longitudinal demand force and yaw moment. The target torques of the first and second drive motors are calculated based on the dynamic feasible domain to obtain the target torque command set.

[0016] Based on the target torque command set, the braking and driving requirements of the corresponding motor are decomposed, the target driving torque or target braking torque is calculated, and the first torque command is obtained.

[0017] Specifically, the analysis of the first state vector constructs state constraints, performs coupled analysis of the vehicle's longitudinal and lateral dynamics, calculates the vehicle's longitudinal demand force and yaw moment, and calculates the dynamic feasible region for the joint operation of the two motors based on the state constraints, including:

[0018] Based on the analysis of the first state vector, the expected longitudinal acceleration is obtained, and the expected lateral turning angle is obtained by analyzing the tunnel curvature information.

[0019] Based on the real-time vehicle status and roadway conditions, calculate the longitudinal demand force and yaw moment corresponding to the expected longitudinal acceleration and expected lateral turning angle, respectively;

[0020] Calculate the maximum lateral force that prevents the vehicle tires from slipping laterally, and combine the maximum lateral force with the maximum longitudinal force of the two motors to construct state constraints;

[0021] Based on the state constraints, the dynamic feasible region for the joint operation of the two motors is calculated.

[0022] Specifically, the step of constructing target points based on longitudinal demand force and yaw moment, and calculating target torques for the first and second drive motors respectively in conjunction with the dynamic feasible region, yields a target torque command set, including:

[0023] Construct the target point based on the longitudinal demand force and the yaw moment;

[0024] Within the dynamic feasible domain, the target torque of the first drive motor and the second drive motor are calculated according to the position of the target point to obtain a set of target torque commands.

[0025] If the target point is not within the dynamic feasible region, the target point is projected onto the dynamic feasible region to obtain the projection point. The target torque of the first drive motor and the second drive motor are calculated according to the position of the projection point to obtain the target torque command set.

[0026] Specifically, the step of decomposing the braking and driving requirements of the corresponding motor according to the target torque command set, calculating the target driving torque or target braking torque, and obtaining the first torque command includes:

[0027] The target torque with a negative value is selected from the target torque command set, the braking demand is analyzed, and the braking demand is decomposed into the first braking part and the second braking part according to the real-time status of the motor.

[0028] Based on the real-time status of the motors, the first braking component is dynamically allocated to each motor, the corresponding target braking torque is calculated, and the second braking component is allocated to the vehicle's hydraulic control component.

[0029] Based on the positive target torque, analyze the drive demand and calculate the target drive torque of the motor;

[0030] The first torque command is obtained by combining the target driving torque or the target braking torque.

[0031] Specifically, the step of sending the status data of the first and second drive motors to the corresponding motor controller based on the dual-motor status data, comparing the speeds of the two motors and calculating the corresponding compensation torque, and optimizing the motor control process according to the compensation torque includes:

[0032] Based on the dual motor status data, the status data of the first drive motor and the second drive motor are sent to the motor controller on the other side respectively;

[0033] The speed difference between the two motors is calculated by comparing their speeds, the cause of the speed error is analyzed and the corresponding compensation torque is calculated, and the first compensation torque command and the second compensation torque command are obtained.

[0034] The motor control process is optimized according to the first compensation torque command and the second compensation torque command.

[0035] Specifically, the step of comparing the speeds of the two motors to calculate the real-time speed difference, analyzing the causes of the speed error and calculating the corresponding compensation torque, and obtaining the first compensation torque command and the second compensation torque command, includes:

[0036] The speed difference between the two motors is calculated by comparing their speeds in real time. The causes of the speed error are analyzed and the corresponding weights are calculated. The causes of the speed error include load differences, electromechanical characteristic differences and tire slippage.

[0037] The load compensation torque for load difference, the efficiency compensation torque for electromechanical characteristic difference, and the slip compensation torque for tire slip are calculated, and then weighted and summed according to their respective weights to obtain the first compensation torque command and the second compensation torque command.

[0038] Specifically, optimizing the motor control process according to the first compensation torque command and the second compensation torque command includes:

[0039] The first compensation torque command and the second compensation torque command are respectively superimposed on the first torque command to obtain the second torque command;

[0040] The motor control process is adjusted and optimized in real time according to the second torque command.

[0041] A dual-motor cooperative control system for a dual-driver trackless rubber-tired vehicle used in mining, used to implement the aforementioned dual-motor cooperative control method for a dual-driver trackless rubber-tired vehicle used in mining, comprising:

[0042] The status analysis module collects real-time operating status data, driver operation data, and roadway environment data of the dual-driver trackless rubber-tired vehicle for mining, analyzes the real-time status of the vehicle, and constructs the first state vector.

[0043] The control command generation module configures a dual-motor cooperative control mechanism, analyzes the first state vector to construct state constraints, performs coupled analysis on the longitudinal and lateral forces of the vehicle, calculates the target torques of the first drive motor and the second drive motor respectively, and obtains the first torque command, wherein the target torque includes the target drive torque and the target braking torque.

[0044] The dual-motor collaborative control module, in response to the first torque command, controls the working process of the first drive motor and the second drive motor respectively, and collects the motor status data in real time to obtain dual-motor status data.

[0045] The control optimization module sends the status data of the first drive motor and the second drive motor to the corresponding motor controller based on the dual motor status data. It compares the speeds of the two motors and calculates the corresponding compensation torque. The control process of the motor is optimized according to the compensation torque to achieve coordinated control of the two motors.

[0046] The beneficial effects of this application are as follows: By integrating and analyzing operational status data, driver operation data, and roadway environment data, and through coupled analysis of the vehicle's longitudinal and lateral demands, the longitudinal demand force and yaw moment are calculated. Combined with state constraints, the dynamic feasible domain of torque distribution between the two motors is calculated. By combining the driver's intention and vehicle status with the dynamic feasible domain, the optimal torque distribution scheme that is closest to the driver's intention and can be safely executed is selected. The braking demand is decomposed into electric braking and hydraulic braking, and electric braking is used first for energy recovery. The electric motor power is dynamically distributed between the two motors, which can optimize braking stability and recovery efficiency. By comparing the motor status data with the real-time speed difference, and by analyzing the error causes such as load differences, electromechanical characteristic differences, and tire slippage, the weighted compensation torque is calculated, and the working status of the two motors is adjusted in real time. This can effectively suppress the cooperative deviation caused by asymmetric factors and improve the system robustness. By optimizing torque distribution and combining efficient braking energy recovery, the driving stability and controllability of the vehicle under complex working conditions can be improved, the risk of instability such as sideslip and fishtailing can be reduced, and the safety of personnel and equipment underground can be ensured. Attached Figure Description

[0047] Figure 1 This is an overall structural diagram of the dual-driver trackless rubber-tired mining vehicle in Embodiment 1 of this application;

[0048] Figure 2 This is a flowchart illustrating the dual-motor coordinated control method for a dual-driver trackless rubber-tired vehicle used in mining, as described in Embodiment 2 of this application.

[0049] Figure 3 This is a flowchart illustrating the calculation process of the first torque command in Embodiment 2 of this application;

[0050] Figure 4This is a schematic diagram of the dual-motor cooperative control system of a dual-driver trackless rubber-tired vehicle for mining, as shown in Embodiment 2 of this application. Detailed Implementation

[0051] The present application will be further described in detail below with reference to the accompanying drawings and embodiments.

[0052] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0053] Hereinafter, the terms "first," "second," and other generic terms are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0054] Example 1:

[0055] Currently, there is a lack of complete sets of transportation equipment for tunneling faces. Auxiliary transportation equipment widely used in mines, such as monorail cranes and tracked vehicles, are limited by the roadway cross-section and drilling conditions, making them unable to reach the tunneling face and meet current needs. Furthermore, monorail cranes and similar equipment are difficult to deploy, generally used for transporting large equipment, have low levels of automation, and are relatively tall, making them inconvenient for carrying personnel and taking up space. The mine-use dual-driver trackless rubber-tired vehicle is simple in design, lightweight, relatively narrow in width, highly maneuverable, has even power distribution, good operability, and integrates video monitoring, data acquisition, data processing, and early warning processing. It is highly expandable, with reserved interfaces, allowing direct uploading of data from long-distance auxiliary transportation systems at the tunneling face according to usage requirements.

[0056] In actual operation, workers need to walk 1.5 to 2 kilometers, which takes 40 to 50 minutes to reach the tunneling face. At the same time, workers also need to carry tools and materials. Due to the limited space in the tunnel and the presence of drilling equipment, belt conveyors and other large machinery along the way, the working environment is complex and crowded. The cross-operation of people and machines not only significantly increases the labor intensity, but also leads to low transportation efficiency and certain safety hazards.

[0057] Figure 1This is a dual-driver trackless rubber-wheeled vehicle for mining, primarily used for transporting personnel and small tools. The vehicle body measures 4025×1200×1800mm and weighs 1800kg. It has two driver's seats (front and rear), two explosion-proof lithium batteries, two explosion-proof control boxes, and two explosion-proof motors. Each battery powers one motor, which in turn controls the motor's rotation via the control cabinet, driving the vehicle. The two motors are synchronized, and the two control cabinets are connected. Control units within the two cabinets coordinate operations such as throttle input, braking, and starting. One motor powers both wheels, and both motors operate simultaneously during vehicle movement. Braking is achieved using a hydraulic wet brake. Real-time parameters of the lights, camera, laser rangefinder, and battery status are displayed on the monitor.

[0058] Preferably, the mining dual-driver trackless rubber-wheeled vehicle can seat 6 people; the seats are removable for convenient transportation of materials and tools, and it has a towing function (it can be towed when there is no power). It uses a right-angle reducer and gear transmission, with gears made of 40Cr tempered steel. The power system uses an explosion-proof lithium battery to power the motor (72V, 230Ah), with a motor power of 7.5KW and an unloaded running speed ≤6.9 m / s. The vehicle's dual motors are controlled simultaneously. Traditional methods use a one-to-one control method, which prevents the vehicle from running normally and easily damages the transmission structure. However, through dual-motor coordinated control, the battery wear is synchronized.

[0059] Due to the special nature of mining equipment, the equipment is generally large in size and weight. For example, monorail cranes and railcars start at tens of tons and are generally used in relatively wide roadways, mainly for transporting large materials. Mining dual-driver trackless rubber-tired vehicles are designed to be compact and are mainly used in narrower working environments. The vehicle weight does not exceed two tons, and there is no need to lay tracks, which overcomes the problems of weight and operating environment. They are equipped with communication interfaces, allowing the mine to collect and manage data in a unified manner. They are simple and quick to set up, easy to install, easy to operate, and highly flexible. Through electrification and intelligent technology, they solve the problems of low transportation efficiency and many safety hazards in narrow roadways. They support multi-mode control and efficient energy management, and can work in conjunction with intelligent mining equipment to improve the overall operational efficiency of the mine.

[0060] Example 2:

[0061] refer to Figure 2 The image shows a specific implementation of a dual-motor cooperative control method for a dual-driver trackless rubber-tired vehicle used in mining, as described in this application, including:

[0062] S101. Real-time acquisition of operating status data, driver operation data and roadway environment data of dual-driver trackless rubber-tired vehicles for mining, analysis of vehicle real-time status, and construction of the first state vector;

[0063] S102. Configure a dual-motor cooperative control mechanism, analyze the first state vector to construct state constraints, perform coupled analysis on the longitudinal and lateral forces of the vehicle, calculate the target torques of the first drive motor and the second drive motor respectively, and obtain the first torque command, wherein the target torque includes the target drive torque and the target braking torque.

[0064] S103. In response to the first torque command, control the working process of the first drive motor and the second drive motor respectively, and collect the status data of the motors in real time to obtain the status data of the two motors.

[0065] S104. Based on the dual motor status data, the status data of the first drive motor and the second drive motor are sent to the opposite motor controller respectively. The speeds of the two motors are compared and the corresponding compensation torque is calculated. The motor control process is optimized according to the compensation torque so as to perform coordinated control of the dual motors.

[0066] In this embodiment, data is collected in real time through a sensor network deployed on the vehicle, including the operating status data, driver operation data, and tunnel environment data of the dual-driver trackless rubber-tired mining vehicle. The operating status data includes, but is not limited to, the rotational speeds of the non-drive and drive wheels collected by wheel speed sensors, the longitudinal acceleration, lateral acceleration, and yaw rate of the vehicle collected by the inertial measurement unit, and the motor speed and temperature collected by the motor controller. The driver operation data includes, but is not limited to, the accelerator pedal opening and brake pedal travel collected by the pedal position sensor, and the steering wheel angle and turning rate collected by the steering wheel angle sensor. The tunnel environment data includes, but is not limited to, the relative distance and speed of obstacles ahead collected by the front-mounted lidar or millimeter-wave radar. The vehicle positioning system, combined with a pre-stored high-precision tunnel map, calculates the vehicle's global coordinates, heading angle, and the geometric parameters of the tunnel in real time.

[0067] Preferably, the system performs time synchronization and spatial alignment processing on the operating status data, driver operation data, and tunnel environment data, and analyzes the real-time vehicle status to construct a first state vector. The real-time status includes, but is not limited to, longitudinal vehicle speed, yaw rate, expected longitudinal acceleration, steering wheel angle, and steering rate. Spatiotemporal alignment of vehicle status, driver input, and tunnel environment information avoids control decision deviations caused by data asynchrony or coordinate system inconsistencies. The constructed first state vector includes all the key information required for vehicle dynamic control, providing accurate data support for subsequent analysis. By fusing pre-stored maps with real-time positioning to obtain tunnel geometric parameters, the control strategy can anticipate curves or slopes ahead, improving stability and safety when driving in complex tunnels.

[0068] Specifically, a dual-motor cooperative control mechanism is configured, and state constraints are constructed by analyzing the first state vector: Based on the tire friction circle and the first state vector, the maximum limit of the resultant force between the tire and the ground is analyzed, and state constraints are constructed by constraining the actual tire force to not exceed the maximum adhesion force; the longitudinal and lateral forces of the vehicle are coupled and analyzed to calculate the total longitudinal force required to meet the driver's longitudinal acceleration expectation and the additional yaw moment required to maintain stable vehicle steering; the dynamic feasible region for the joint operation of the two motors is calculated, using the torques of the two motors as control variables and the state constraints as boundaries to construct the dynamic feasible region. Using the calculated total longitudinal force and additional yaw moment as objectives, the optimal solution is sought within the dynamic feasible region, the target torque is calculated separately, and the longitudinal force required by the left and right drive wheels is solved. Then, based on the wheel radius and reduction ratio, the target drive torque or target braking torque of the left and right motors is converted to obtain the first torque command.

[0069] It's important to note that by constructing a dynamic feasible region based on the tire friction circle and motor capabilities in real time, it's possible to ensure that control commands do not cause vehicle instability or exceed the actuator's capabilities, thus improving the safety of the control process. This avoids the problems of conflict or abrupt switching between longitudinal drive and lateral stability control found in traditional control systems. For example, during cornering acceleration, more torque can be automatically allocated to the outer motor, providing both acceleration and yaw moment conducive to stable steering, rather than a simple average distribution. Combining this with the dynamic feasible region selection to find the optimal solution aligns with the driver's intended trends and ensures vehicle controllability.

[0070] Furthermore, the vehicle domain controller or vehicle controller sends a first torque command containing the target torque values ​​for the left and right motors to the left and right motor controllers respectively via the CAN bus. Upon receiving its corresponding torque command, each controller uses a high-precision rotary transformer to collect the actual motor speed and rotor position angle in real time, and uses a current sensor to collect the three-phase current of the motor, obtaining dual-motor status data. The controller uses a vector control algorithm to control the motors. The vector control algorithm combines the first torque command with the current motor speed and DC bus voltage, and through online calculation, determines the desired d-axis and q-axis current reference values ​​under the current operating conditions. The d-axis current is used for field weakening, and the q-axis current is used to generate torque. The detected three-phase AC current is transformed into the actual d-axis and q-axis currents in a synchronous rotating coordinate system using Park and Clarke transformations. A PI controller is used to adjust the d-axis and q-axis current errors, outputting the corresponding d-axis and q-axis voltage reference values. Through inverse Park transformation and space vector pulse width modulation, a PWM wave signal is generated to drive the inverter power switch, controlling the motor output to match the actual electromagnetic torque of the first torque command. Throughout the process, the controller continuously collects and updates the motor's status data, including but not limited to motor winding temperature, controller temperature, and DC bus voltage. The data is fed back to the upper-level controller in real time via the CAN bus for status monitoring and subsequent optimization.

[0071] It is important to emphasize that FOC-based motor control enables decoupled torque control, allowing the motor to quickly and accurately track torque commands from the upper layer across a wide speed range. It offers fast response, real-time acquisition of motor speed to determine synchronization between motors and whether slippage occurs on one side, and motor current and temperature data to reflect the actual load and thermal state of the motor, which can be used for overload protection and health management. Through closed-loop current and speed control, combined with temperature and voltage monitoring, it can effectively prevent motor overcurrent, overheating, overvoltage, or step loss faults, improving the robustness and lifespan of the drive system.

[0072] Specifically, the left and right motor controllers establish direct communication via a high-speed CAN bus. Each controller periodically sends its core status data to the controller on the other side. Each controller, referred to as an MCU, obtains the real-time status of its own and the other side's motors. The controller compares the speeds of the two motors and calculates the corresponding compensation torque. The controller calculates the real-time speed difference between the left and right motors and analyzes the causes. These causes include load differences due to uneven distribution of vehicle load or greater road resistance on one side, electromechanical characteristic differences caused by slight inconsistencies in the inverter characteristics of the motors or sensor zero bias, and dynamic tire slippage caused by one tire entering the nonlinear slip zone. The corresponding compensation torque is calculated for different causes and fused to generate a first compensation torque command and a second compensation torque command acting on the left and right motors respectively. The calculated compensation torque commands are superimposed on the first torque command in real time to obtain the second torque command. The motor control process is optimized based on the second torque command.

[0073] It should be noted that through real-time comparison and compensation, systematic errors caused by manufacturing tolerances, aging, and asymmetrical loads can be effectively offset, ensuring that the actual driving or braking force of the left and right wheels is strictly distributed as needed, avoiding deviation or ineffective energy loss caused by long-term operation; when one side of the vehicle suddenly encounters a low-traction road surface, the slip compensation mechanism based on the speed difference can quickly intervene to suppress slippage and enhance the system's robustness to road disturbances and sudden operating conditions; through statistical analysis of long-term compensation data, subtle changes in motor characteristics can be identified in reverse, compensation parameters can be updated, closed-loop optimization can be performed, and collaborative optimization control of the two motors can be achieved.

[0074] This application integrates and analyzes operational status data, driver operation data, and roadway environment data. By coupling the longitudinal and lateral demands of the vehicle, it calculates the longitudinal demand force and yaw moment, and, combined with state constraints, calculates the dynamic feasible region for torque distribution between the two motors. Combining driver intent and vehicle status with the dynamic feasible region, it selects the optimal torque distribution scheme that best approximates the driver's intent and is safe to execute. Braking demand is decomposed into electric braking and hydraulic braking, prioritizing energy recovery through electric braking. Dynamically distributing electric motor power between the two motors optimizes braking stability and recovery efficiency. By comparing motor status data with real-time speed differences and analyzing errors such as load differences, electromechanical characteristic differences, and tire slippage, it calculates weighted compensation torque and adjusts the dual-motor operating status in real time. This effectively suppresses collaborative deviations caused by asymmetric factors and improves system robustness. By optimizing torque distribution and combining it with efficient braking energy recovery, it enhances the vehicle's driving stability and controllability under complex operating conditions, reduces the risk of instability such as sideslip and fishtailing, and ensures the safety of personnel and equipment underground.

[0075] Furthermore, real-time data collection of the operating status, driver operation data, and roadway environment data of the dual-driver trackless rubber-tired mining vehicle is conducted to analyze the vehicle's real-time status and construct a first state vector, including:

[0076] S201. Real-time acquisition of operating status data, driver operation data, and roadway environment data of dual-driver trackless rubber-tired vehicles for mining, to obtain multi-source vehicle data;

[0077] S202. Perform time synchronization and spatial alignment processing on multi-source vehicle data, and analyze the geometric parameters of the roadway, vehicle pose parameters and environmental obstacle parameters to construct the first state vector.

[0078] In this embodiment, real-time data acquisition is performed on the operating status of the dual-driver trackless rubber-tired mining vehicle, including driver operation data and tunnel environment data. Operating status data includes, but is not limited to, the rotational angular velocity, linear acceleration, and angular velocity of each wheel; the real-time speed, torque current, and winding temperature of the motor. Specifically, a magnetoelectric or Hall effect wheel speed sensor is installed at each wheel hub to measure the rotational angular velocity of each wheel; a six-axis inertial measurement unit is installed near the vehicle's center of gravity to measure the linear acceleration and angular velocity of the vehicle along its three axes; a rotary transformer is installed at the output shaft of the drive motor or at the reducer to obtain the real-time speed, torque current, and winding temperature of the motor through the motor controller. Driver operation data includes, but is not limited to, the driver's demand for longitudinal power intensity and direction, and the driver's lateral control intentions. Specifically, linear or rotary potentiometers are installed at the pivots of the accelerator and brake pedals to convert their mechanical travel into voltage signals, reflecting the driver's demand for longitudinal power intensity and direction; an absolute encoder is installed on the steering column to measure the steering wheel angle and its rotation rate, reflecting the driver's lateral control intentions.

[0079] Specifically, the tunnel environment data includes, but is not limited to, the distance to obstacles ahead. Specifically, a multi-line LiDAR is installed above the vehicle's front bumper. Its emitted laser beam scans a fan-shaped area ahead, obtaining point cloud data on the distance, size, and relative velocity of obstacles by measuring the beam reflection time. The vehicle's onboard positioning module calculates the vehicle's three-dimensional coordinates and heading angle in the tunnel's global coordinate system in real time. This positioning information is then matched against a pre-stored high-precision digital map of the tunnel in the onboard controller to obtain geometric constraints such as the radius of curvature, slope angle, and width of the tunnel segment where the vehicle is currently located. Analog or digital signals collected by sensors are connected to the vehicle's domain controller via their respective data lines. The controller's internal acquisition program periodically reads, converts, and caches these signals, for example, at 10-millisecond intervals, to obtain multi-source vehicle data.

[0080] It should be noted that by simultaneously acquiring wheel speed and yaw rate, the actual steering state of the vehicle can be cross-verified; by acquiring steering wheel angle and roadway curvature, it is possible to analyze whether the driver's steering intention matches the environment; and by collecting motor speed and steering wheel angle, accurate data support can be provided for subsequent torque distribution and steering compensation.

[0081] Furthermore, time synchronization and spatial alignment processing are performed on multi-source vehicle data. For directly accessed analog signals, a timestamp is recorded the instant the A / D conversion is completed. At the beginning of each control cycle, based on the minimum time deviation interpolation algorithm, all data to be used are uniformly interpolated to the same reference time point, such as the start of the current cycle, ensuring that the acceleration, wheel speed, steering wheel angle, etc., used for calculation are in the same state at the same moment. Through a pre-calibrated rigid transformation matrix between the radar and the vehicle's center of gravity, the coordinates of each obstacle point are transformed into a vehicle coordinate system with the vehicle's center of gravity as the origin, the front direction as the X-axis, the left side as the Y-axis, and the upward direction as the Z-axis. The global coordinates from the positioning module are transformed into lateral offsets relative to the current lane centerline through coordinate transformation. The geometric parameters of the lane, vehicle pose parameters, and environmental obstacle parameters are analyzed to construct the first state vector.

[0082] Specifically, by querying a pre-stored digital map, the vehicle obtains the centerline curvature sequence and slope sequence of the alleyway within a predetermined distance (e.g., 50 meters) ahead, based on its current position. These sequences are simplified into representative values ​​for control decisions, such as the radius of curvature and slope angle of the nearest curve ahead, yielding the alleyway's geometric parameters. Combining synchronized IMU data (heading angular velocity) and wheel speed data, Kalman filtering is used to estimate the vehicle's yaw angle, longitudinal velocity, lateral velocity, and lateral distance and heading angle deviation relative to the alleyway's centerline, resulting in the vehicle's pose parameters. The LiDAR point cloud, transformed to the vehicle coordinate system, is clustered to identify the nearest obstacle in the same lane ahead. Its longitudinal distance and relative velocity relative to the vehicle are calculated, along with the estimated collision time, yielding environmental obstacle parameters. The calculated alleyway geometric parameters, vehicle pose parameters, and environmental obstacle parameters are then combined sequentially to obtain the first state vector.

[0083] It should be noted that time synchronization ensures that control decisions are based on vehicle data at the same moment, avoiding conflicting control commands caused by data delays and asynchrony. Spatial alignment unifies all environmental information into a vehicle-centric decision coordinate system, providing a unified geometric basis for subsequent calculations of tire forces, yaw moments, etc. By extracting state parameters and constructing a first state vector, the real-time state of the vehicle's operation can be reflected, providing accurate data support for subsequent motor control processes and reducing the computational complexity and design difficulty of the control process.

[0084] Furthermore, the dual-motor cooperative control mechanism includes:

[0085] S301. Analyze the first state vector to construct state constraints, perform coupled analysis on the longitudinal and lateral forces of the vehicle, calculate the longitudinal demand force and yaw moment of the vehicle, and calculate the dynamic feasible region of the dual motors working together based on the state constraints.

[0086] S302. Construct target points according to longitudinal demand force and yaw moment, and calculate the target torques of the first drive motor and the second drive motor respectively in combination with the dynamic feasible domain to obtain the target torque command set.

[0087] S303. Based on the target torque command set, decompose the braking and driving requirements of the corresponding motor, calculate the target driving torque or target braking torque, and obtain the first torque command.

[0088] In this embodiment, the first state vector is analyzed to construct state constraints, and the longitudinal and lateral forces of the vehicle are coupled and analyzed to calculate the longitudinal demand force and yaw moment of the vehicle. Based on the state constraints, the dynamic feasible region for the joint operation of the two motors is calculated. By calculating the feasible region based on the physical limits of the tires and the capabilities of the motors, a safe range of control commands is determined, ensuring that no subsequent allocation scheme will lead to vehicle instability or actuator saturation. By decoupling the longitudinal and lateral control objectives and optimizing based on the feasible region, the dynamic feasible region can be updated according to the real-time state, enabling the control strategy to adaptively adjust the expectations.

[0089] The target point is constructed based on the longitudinal demand force and yaw moment. The target torque of the first drive motor and the second drive motor are calculated separately in combination with the dynamic feasible region to obtain the target torque command set. When the target point is within the dynamic feasible region, the control system can execute accurately without loss or delay. When the target point is not within the dynamic feasible region, it can be automatically and smoothly adjusted to the nearest safe point through the projection method. Under the premise of ensuring stability, the driver's driving intention is preserved to the greatest extent. For example, at the limit of the curve, the driving force will be slightly reduced but most of the steering torque will be retained instead of completely abandoning the steering response. The target torque of the first drive motor and the second drive motor are calculated separately. The driving or braking torque of the left and right wheels can be actively and independently controlled to obtain accurate yaw moment, which can improve the flexibility and stability of the handling.

[0090] Specifically, based on the target torque command set, the braking and driving requirements of the corresponding motors are decomposed, and the target driving torque or target braking torque is calculated to obtain the first torque command. Regenerative braking is prioritized, and the regenerative braking force is distributed between the two motors. This maximizes the recovery of kinetic energy during vehicle braking, converting it into electrical energy stored in the battery, reducing energy consumption in underground roadway conditions with frequent start-stop operations. By coordinating the timing and force of intervention of motor braking and hydraulic braking, compound braking can be performed, avoiding sudden changes in braking force and improving the smoothness and stability of the braking process. Dynamically distributing the regenerative braking force of the left and right wheels according to the vehicle status helps maintain directional stability during braking, prevents braking deviation on uneven road surfaces, and improves the robustness of the system.

[0091] Furthermore, the first state vector is analyzed to construct state constraints, and a coupled analysis of the vehicle's longitudinal and lateral dynamics is performed to calculate the vehicle's longitudinal demand force and yaw moment. Based on the state constraints, the dynamic feasible region for the joint operation of the two motors is calculated, including:

[0092] S401. Analyze the longitudinal acceleration expectation based on the first state vector, and analyze the tunnel curvature information to obtain the lateral turning angle expectation;

[0093] S402. Based on the real-time vehicle status and roadway conditions, calculate the longitudinal demand force and yaw moment corresponding to the expected longitudinal acceleration and expected lateral turning angle, respectively.

[0094] S403. Calculate the maximum lateral force that prevents the vehicle tires from slipping laterally, and construct state constraints by combining the maximum lateral force and the maximum longitudinal force of the two motors.

[0095] S404. Calculate the dynamic feasible region for the joint operation of the two motors based on the state constraints.

[0096] In this embodiment, based on the analysis of the longitudinal acceleration expectation using the first state vector, the controller extracts the accelerator pedal opening signal and brake pedal travel signal from the driver's operation data from the first state vector. After low-pass filtering to eliminate high-frequency noise and pedal jitter, the two analog signals are input into a pre-calibrated one-dimensional lookup table to calculate the vehicle's expected longitudinal acceleration value under the current operation. Positive values ​​correspond to acceleration, and negative values ​​correspond to deceleration. The calibration relationship of the one-dimensional lookup table is determined through vehicle testing. The input is the processed pedal signal, and the output is the vehicle's expected longitudinal acceleration value under the current operation. Simultaneously, the road slope information extracted from the lane environment data is added as feedforward compensation. When going uphill or downhill, the same pedal opening can generate the expected acceleration required to maintain the target vehicle speed, thus realizing the hill start assist function.

[0097] Simultaneously, the expected lateral turning angle is obtained by analyzing the tunnel curvature information based on the first state vector. The driver's steering wheel angle, directly measured by the steering wheel angle sensor, is obtained from the first state vector. From the pre-stored high-precision tunnel map, based on the vehicle's current position and heading, the curvature of the tunnel centerline ahead is extracted. The tunnel centerline curvature is then converted into a reference front wheel angle that aligns the vehicle trajectory with the future tunnel centerline using Ackerman steering geometry. When the driver does not significantly turn the steering wheel, the reference front wheel angle calculated from the tunnel centerline curvature is used as the primary expectation for path following assistance. When the driver actively applies steering input, the process gradually switches to driver intent as the primary factor, combined with tunnel curvature for boundary safety correction. The expected lateral turning angle is obtained by fusing the tunnel curvature information and the first state vector, where the fusion weight is calculated based on the steering wheel angle change rate.

[0098] It should be noted that, through calibration mapping and filtering, the driver's pedal and steering operations are converted into continuous and standard acceleration and angle expectation signals, eliminating noise and uncertainty in the operation signals and providing accurate data input for subsequent control processes. By combining the lateral expectation with the curvature analysis of the roadway centerline, the resulting vehicle control target not only responds to the current driver operation but also takes into account the roadway direction ahead. For example, before entering a curve, even if the driver has not yet turned the steering wheel, a slight angle expectation will be generated to guide the vehicle into the curve smoothly. This can improve the smoothness and trajectory accuracy of driving in the roadway, reduce the driver's lateral control burden, and improve the accuracy of the control process.

[0099] Specifically, based on the real-time vehicle status and the roadway conditions, the longitudinal demand force and yaw moment corresponding to the expected longitudinal acceleration and expected lateral turning angle are calculated respectively. Specifically, the calculation is performed through a preset vehicle dynamics model, which adopts a linear two-degree-of-freedom vehicle model. The model input includes: the expected longitudinal acceleration and expected lateral turning angle, as well as the real-time vehicle status in the first state vector, including longitudinal speed, vehicle mass, vehicle moment of inertia about the vertical axis, wheelbase, and distance from the center of mass to the front and rear axles. The model calculates and outputs the longitudinal demand force and yaw moment.

[0100] The specific calculation process of the vehicle dynamics model includes: calculating the longitudinal force required by the vehicle to overcome inertial force, rolling resistance and gravity component of the slope; determining the ideal yaw rate by analyzing the expected lateral turning angle and vehicle speed; adjusting the deviation between the ideal yaw rate and the actual measured yaw rate to obtain the yaw moment.

[0101] Specifically, the formula for calculating vertical demand force is:

[0102] ;

[0103] In the formula, For vertical demand forces, For vehicle quality, For the expected longitudinal acceleration, The slope angle of the tunnel. It is the acceleration due to gravity. The rolling resistance coefficient is calculated by superimposing the vehicle's inertial force, rolling resistance, and the gravity component of the ramp to determine the longitudinal force required to generate the desired acceleration. The rolling resistance coefficient is typically determined through bench tests or real-vehicle coasting tests. In bench tests, the vehicle's drive wheels are placed on a rotating test bench, and rolling resistance is measured under different loads and speeds. Real-vehicle coasting tests involve allowing the vehicle to coast from an initial velocity in neutral on a flat road surface, and the rolling resistance coefficient is inferred by measuring the deceleration. When setting this coefficient, an empirical range is selected based on common road surface types in mining areas (such as dry hardened tunnels, wet muddy roads, and gravel-paved roads), or a standard reference value can be provided by the vehicle manufacturer.

[0104] Specifically, the ideal yaw rate is calculated from the expected lateral turning angle and vehicle speed using a linear two-degree-of-freedom model formula. The calculation formula is as follows:

[0105] ;

[0106] In the formula, For the ideal yaw rate, For the longitudinal speed of the vehicle, For the expected lateral turning angle, Wheelbase The understeer gradient coefficient is determined through steady-state turning tests on real vehicles (such as accelerated cornering with a fixed steering wheel angle): the difference in front and rear axle slip angles under different lateral accelerations is measured, and the actual understeer gradient is derived. The setting is based on ensuring that the linear two-degree-of-freedom model accurately reflects the inherent steering response characteristics of the vehicle. A larger K value results in a smaller ideal yaw rate predicted by the model, and the yaw moment required by the controller to compensate for the deviation is adjusted accordingly, thereby achieving precise and stable control of vehicles with different steering characteristics. The yaw moment is obtained by adjusting the deviation between the ideal yaw rate and the actual measured yaw rate using a proportional-derivative (PD) controller. The specific calculation formula is as follows:

[0107] ;

[0108] In the formula, For yaw moment, This is the proportionality coefficient. The differential coefficients are... This is the actual yaw rate. For time. Yaw moment is used to eliminate the difference between the actual yaw response and the ideal model response, suppressing oversteer or understeer.

[0109] It should be noted that by calculating the longitudinal demand force and yaw moment based on the driver's motion commands using the vehicle dynamics model, the yaw moment can be continuously fine-tuned during steering without significant instability to optimize the vehicle's steering characteristics and make them more in line with the ideal response. The calculated longitudinal demand force and yaw moment provide accurate data support for the subsequent determination of target points. The longitudinal demand force and yaw moment calculated based on the same time and the same state are coupled with the longitudinal speed demand and the lateral stability correction demand, which can improve the accuracy and precision of tire force distribution.

[0110] Specifically, the maximum lateral force required to prevent lateral slippage of the vehicle tires is calculated. For each tire, the current vertical load is determined based on the static load and the dynamic load transfer caused by the vehicle's longitudinal and lateral accelerations. The dynamic load transfer is calculated using the longitudinal / lateral accelerations in the first state vector and the vehicle's geometric parameters. The maximum adhesion force is obtained by multiplying the current vertical load of the tire by the tire-road adhesion coefficient in the first state vector. The tire-road adhesion coefficient is a dimensionless parameter representing the ratio of the maximum friction force that can be generated between the tire and the road surface to the tire's vertical load. This coefficient is obtained through an online estimation algorithm or a preset road surface mode (selectable by the driver or based on environmental perception). The setting is based on ensuring that the state constraints accurately reflect the current road surface limits, preventing control commands from exceeding the tire's grip safety boundary, thereby actively limiting driving force or intervening in yaw moment on low-adhesion surfaces to prevent vehicle sideslip. According to the friction circle theory, when longitudinal and lateral forces are present simultaneously, the following inequality must be satisfied:

[0111] ;

[0112] In the formula, For longitudinal force, It is a lateral force. To achieve maximum adhesion, the absolute value of the longitudinal force on each tire is limited to no more than [a certain value]. , represented as:

[0113] ;

[0114] In the formula, A safety factor less than 1 can be set according to the vehicle control accuracy requirements, for example, 0.9, using the inequality as a tire adhesion constraint. The maximum longitudinal force of the motor is obtained. Based on the maximum output torque of each drive motor, combined with the reduction ratio and wheel radius, the maximum output torque is calculated by multiplying the reduction ratio by the wheel radius to obtain the maximum longitudinal force. This maximum output torque is then converted into the maximum longitudinal force at the wheel. State constraints are constructed for each drive wheel, namely: In the formula, For the maximum longitudinal force, the state constraints clearly define the upper and lower limits of the safe and effective longitudinal force that each motor can currently provide.

[0115] It should be noted that by using constraints based on the tire friction circle, the longitudinal force allocated to the drive wheels is ensured to not sacrifice necessary lateral adhesion, which can effectively prevent sideslip or fishtailing caused by the drive wheels exceeding the adhesion limit, thus improving the stability of the vehicle control process. Combined with motor characteristic constraints, the control commands are ensured to always be within the physical operating capacity of the motor, avoiding invalid commands or situations that lead to motor overheating or overload damage. By constructing dynamic state constraints, the state constraints can be updated in real time according to load transfer, changes in adhesion coefficient, and changes in motor speed, so that optimal control can always be performed according to physical limits during the control process, improving control effect and resource utilization.

[0116] Specifically, the controller treats the longitudinal forces of the left and right drive wheels as two freely distributable control variables subject to state constraints. The total driving force obtained by the vehicle is the sum of these two forces, and the yaw moment directly generated by the difference in driving forces between the left and right wheels is proportional to the difference between these two forces. Since the longitudinal force of each wheel is constrained between a minimum and a maximum value, a closed geometry is defined on the total driving force-yaw moment plane through inequality constraints, serving as the dynamic feasible region. This region is a convex polygon. Its vertices can be obtained by enumerating various combinations of the boundary values ​​of the left and right wheel forces when they take their respective constraints, and calculating the corresponding total driving force and yaw moment. For example, when the left wheel force takes the minimum value and the right wheel force takes the minimum value, a vertex is obtained, corresponding to the maximum total braking force and zero yaw moment; when the left wheel force takes the minimum value and the right wheel force takes the maximum value, another vertex is obtained, corresponding to a specific total driving force and the maximum positive yaw moment. Connecting all the vertices, the current dynamic feasible region is obtained. The shape, size and position of the region move and expand in real time with the changes in motor capability caused by changes in vehicle speed and the changes in tire force constraints caused by changes in load transfer and road surface adhesion.

[0117] It should be noted that by integrating the constraints of individual actuators, a global description of the overall vehicle performance is obtained, enabling the controller to accurately obtain the full range of the vehicle's capabilities in the current environment. The dynamic feasible domain provides operational space for safety optimization decisions. By determining whether the target point is within this region, it can be ensured that the generated torque distribution schemes are feasible and stable, thereby improving decision-making efficiency and safety.

[0118] Furthermore, target points are constructed based on longitudinal demand force and yaw moment. Combined with the dynamic feasible region, target torques for the first and second drive motors are calculated respectively, resulting in a target torque command set, including:

[0119] S501. Construct target points according to longitudinal demand force and yaw moment;

[0120] S502. If the target point is within the dynamic feasible region, calculate the target torque of the first drive motor and the second drive motor according to the position of the target point to obtain the target torque command set.

[0121] S503. If the target point is not within the dynamic feasible region, project the target point onto the dynamic feasible region to obtain the projection point. Calculate the target torque of the first drive motor and the second drive motor according to the position of the projection point to obtain the target torque command set.

[0122] In this embodiment, the controller receives the calculated longitudinal demand force and yaw moment, and normalizes them. The reference value used for normalization can be the maximum driving force that the vehicle can theoretically achieve in the current state or the maximum yaw moment calculated based on tire adhesion limits and wheelbase. The longitudinal demand force and yaw moment are divided by their respective reference values, resulting in dimensionless longitudinal demand force and yaw moment. Based on the normalized longitudinal demand force and yaw moment, points on the total driving force-yaw moment plane are selected as target points, reflecting the comprehensive dynamic state that the vehicle is expected to achieve in the current control cycle. By selecting target points, the input interface and problem complexity of subsequent algorithms can be simplified.

[0123] For a target point within the dynamic feasible region, let the coordinates of the target point represent the desired total driving force and the desired yaw moment. The controller knows the vehicle's wheelbase. According to the principle of vehicle force balance, the total driving force is equal to the sum of the longitudinal forces of the left and right drive wheels. The desired yaw moment is proportional to the difference between the longitudinal forces of the left and right drive wheels multiplied by half the wheelbase. The specific values ​​of the longitudinal forces that the left and right drive wheels should provide can be calculated. After obtaining the longitudinal force requirements of the left and right wheels, the controller converts the longitudinal force requirements on the tires into torque requirements on the motor output shaft based on the known vehicle transmission system parameters, mainly the effective rolling radius of the wheels and the transmission ratio of the main reducer. The calculated torque values ​​of the left and right motors constitute the target torque command set for the current control cycle.

[0124] It should be noted that when the system has sufficient capacity, the actual longitudinal and lateral dynamic response of the vehicle can fully match the ideal expectation, providing the driver with accurate control. The calculation process takes less time, which can improve the real-time performance of the vehicle control system, ensure timely response to rapidly changing driving commands, solve for the different torques of the left and right motors, improve the accuracy of yaw moment, and continuously optimize the vehicle's steering characteristics and driving stability through subtle torque differences during normal driving, thereby improving the vehicle's handling agility.

[0125] If the target point is not within the dynamic feasible region, it is projected onto the dynamic feasible region to obtain a projection point. By projecting onto the boundary or interior of the dynamic feasible region, a projection point is found that minimizes the straight-line distance between the projection point and the target point. For convex polygons, the projection point can be on the boundary line or at a vertex. The coordinates of the projection point are obtained and used as the total driving force and yaw moment after projection. Based on the coordinate values ​​of the projection point, the target longitudinal force of the left and right drive wheels is calculated and then converted into the target torque of the left and right motors, resulting in a target torque command set. The vehicle motion state corresponding to this command set is physically realizable and is the closest to the driver's original intention among all realizable states.

[0126] It should be noted that when the driver operates too aggressively, the system will not directly cut off power or fix a certain parameter. Instead, it will automatically adjust to the closest safe operating point through a projection method. For example, it may appropriately reduce the total driving force while preserving the stability adjustment torque as much as possible. This helps the vehicle to corner stably to the greatest extent possible while preventing slippage and instability. Since the dynamic feasible domain is continuously changing, the projection point of the target point near its boundary is also continuously changing. As the vehicle gradually approaches its physical limits, the actual distribution of motors will show a smooth and abrupt transition. This avoids abrupt changes in torque, maintains vehicle stability, improves ride comfort, and protects transmission components.

[0127] like Figure 3 As shown, based on the target torque command set, the braking and driving requirements of the corresponding motor are decomposed, the target driving torque or target braking torque is calculated, and the first torque command is obtained, including:

[0128] S601. Select the target torque that is negative from the target torque command set, analyze the braking demand, and decompose the braking demand into the first braking part and the second braking part according to the real-time status of the motor.

[0129] S602. Based on the real-time status of the motor, the first braking component is dynamically allocated to each motor, the corresponding target braking torque is calculated, and the second braking component is allocated to the hydraulic control component of the vehicle.

[0130] S603. Based on the positive target torque, analyze the drive demand and calculate the target drive torque of the motor.

[0131] S604. Combine the target driving torque or target braking torque to obtain the first torque command.

[0132] In this embodiment, the controller traverses the target torque command set and identifies commands with negative values, the absolute value of which represents the total torque demanded by the vehicle for the electric motor braking function. The controller analyzes the real-time availability of the regenerative braking system under current conditions, including: obtaining the current rotational speeds of the left and right motors from the motor controller and querying a pre-stored motor external characteristic performance map, which defines the maximum regenerative braking torque that the motor can safely provide as a generator at different speeds; obtaining the real-time status of the power battery from the battery management system, including but not limited to the state of charge and the current maximum acceptable charging power, the charging power limiting the total recovery rate of regenerative braking energy, and calculating the maximum regenerative braking power limit based on the battery status.

[0133] Specifically, the motor external characteristic performance profile describes the boundary curves of the maximum torque that the motor can safely output at different speeds. The construction process includes: mounting the motor under test on a dedicated motor test bench system, which includes a dynamometer, torque sensor, speed sensor, and cooling device; within the motor's allowable operating temperature range, starting from the lowest speed, gradually increasing the motor's load torque until its thermal limit, current limit, or mechanical limit is reached, while simultaneously recording the corresponding maximum steady-state torque value at that speed. For regenerative braking mode, the dynamometer drives the motor to rotate, measuring the maximum regenerative torque at different speeds when the motor is operating as a generator; the maximum torque at each speed point is tested sequentially, and all test points are connected to form a complete drive and regenerative braking external characteristic curve; the curve is stored in the motor controller or vehicle controller, with the motor speed as input and the maximum allowable torque at the corresponding speed as output, used for real-time constraint control commands. The accuracy of this profile affects the accuracy of estimating the regenerative braking torque capability.

[0134] Preferably, the controller combines the motor capacity with the battery limitation, taking the smaller of the two as the maximum regenerative braking torque capacity currently available to the entire vehicle. The total motor braking demand is compared with the maximum regenerative braking torque capacity. If the total demand is less than or equal to the maximum regenerative braking torque capacity, all braking demand can be met by regenerative braking; the first braking portion equals the total demand, and the second braking portion is zero. If the total demand exceeds the maximum regenerative braking torque capacity, the excess cannot be recovered and is provided by the hydraulic braking system; the first braking portion equals the maximum regenerative braking torque capacity, and the second braking portion is the difference between the total demand and the maximum regenerative braking torque capacity.

[0135] It should be noted that by evaluating and prioritizing regenerative braking in real time, the kinetic energy of the vehicle during deceleration is efficiently converted into electrical energy for storage. This is suitable for mining vehicles with frequent start-stop cycles and slow downhill driving, which can reduce overall operating energy consumption, extend electric driving range, and reduce the load on the on-board generator set. Based on the real-time physical limits of the motor's external characteristics and the battery's charging capacity, the system can prevent the risk of motor overspeeding, overheating damage, or battery overcharging caused by excessive pursuit of energy recovery, thereby improving the safety and stability of the system.

[0136] For the first braking component, i.e., the total regenerative braking demand, the controller dynamically allocates the first braking component to each motor, calculates the corresponding target braking torque, and ensures vehicle directional stability during braking while optimizing total energy recovery efficiency. Specifically, vehicle directional stability analysis includes: analyzing the vehicle's real-time state, such as whether there is steering input, whether the yaw rate is stable, and analyzing the differences in road surface adhesion between the left and right sides. By ensuring that uneven braking forces between the left and right sides will not cause unexpected yaw moments that lead to vehicle deviation, the corresponding regenerative braking force ratio between the left and right wheels is calculated. Optimizing total energy recovery efficiency analysis includes: since the speeds of the two motors may differ due to turning or uneven road surfaces, their power generation efficiencies also differ. The motor currently operating in the higher power generation efficiency range is assigned more braking torque to recover more energy while meeting the total braking force demand. Combining the vehicle directional stability analysis and the total energy recovery efficiency optimization analysis, a linear optimization algorithm is used to calculate the target regenerative braking torque values ​​for the left and right motors respectively.

[0137] For the second braking component, namely hydraulic braking, the controller allocates it to the vehicle's hydraulic system. Based on the preset front-to-rear axle braking force distribution curve and considering the real-time vehicle load distribution, it calculates and distributes the target pressure to each wheel's brake cylinder. Upon receiving these pressure commands, the hydraulic control unit precisely controls the brake fluid flow rate via solenoid valves to generate the required hydraulic braking force. The front-to-rear axle braking force distribution curve describes the distribution relationship between the front and rear axles during braking. It is typically presented as the proportion of front axle braking force to total braking force (e.g., the β line) or an ideal braking force distribution curve (I curve). The I curve represents the ideal distribution relationship where the front and rear wheels lock simultaneously, based on parameters such as the vehicle's center of gravity, wheelbase, and track width. In practical engineering, to maintain braking directional stability and prevent the rear wheels from locking up first, a fixed or adjustable distribution curve lower than the I curve is used. In this embodiment, this curve is pre-stored in the controller. The input is the total braking demand or current deceleration, and the output is the target braking force proportion that the front and rear axles should bear. Because the load on mining vehicles varies greatly, the offset of the distribution curve can be dynamically adjusted according to the real-time load distribution, so that the braking force distribution is more in line with the actual wheel load, thereby improving braking stability and shortening the braking distance. The hydraulic control unit calculates the target pressure of each wheel cylinder based on this and then executes it through the hydraulic adjustment unit.

[0138] It should be noted that by considering the vehicle's dynamic state to dynamically distribute regenerative braking force, the yaw moment that may cause instability can be actively suppressed or compensated during braking, which can improve the directional stability and safety of braking on wet, uneven roads or when turning. By combining the differences in motor efficiency for intelligent distribution, the total energy recovery of the system can be increased under the same total regenerative braking demand, and the energy flow management efficiency of the whole vehicle can be optimized.

[0139] Specifically, the controller selects positive values ​​from the target torque command set, with positive values ​​representing the demand for motor output drive torque. The controller continuously reads key status data from the motor controller, including but not limited to the real-time temperature of the motor windings, the power module temperature of the motor controller, and the current speed of the motor. Based on the current speed of the motor, it queries its external characteristic performance spectrum to obtain the maximum allowable output torque capability at that speed. The calculation process includes: analyzing whether the original drive demand exceeds the maximum output torque capability at the current speed; if so, limiting it to that capability value; and using the value after torque capability limiting as the target drive torque.

[0140] It should be noted that by using real-time thermal monitoring and torque derating, it is possible to effectively prevent overheating damage to the motor and controller caused by continuous overload or poor heat dissipation, thereby improving the durability and reliability of the power unit. It ensures that the output of the driving force is always within the physically achievable range, avoiding weak acceleration, system alarms or unpredictable behavior caused by requests exceeding the actuator limits, so that the vehicle's power response is always stable and in line with expectations.

[0141] The calculated target regenerative braking torque (negative value) and target drive torque (positive value) of the left and right motors are integrated. For each motor, only one torque command mode can be executed at any given time, and the current working mode and corresponding final torque value are determined based on the numerical sign. Simultaneously, the controller converts the braking portion that needs to be handled by the hydraulic system into specific hydraulic control commands, including but not limited to the target braking master cylinder pressure or the target pressure values ​​of each wheel cylinder. According to a predefined, highly reliable vehicle network communication protocol, the commands are sent to the corresponding motor controller and hydraulic control unit, and the set of commands from the underlying actuators that operate within the current control cycle is taken as the first torque command.

[0142] It should be noted that by generating and sending commands, the synchronization of the drive motor and the hydraulic braking system can be ensured, improving the anti-interference capability and reliability of key control command transmission in complex vehicle electromagnetic environments, and reducing the risk of system failure caused by communication errors.

[0143] Furthermore, based on the dual-motor status data, the status data of the first and second drive motors are sent to the corresponding motor controller. The speeds of the two motors are compared, and the corresponding compensation torque is calculated. The motor control process is optimized according to the compensation torque, including:

[0144] S701. Based on the dual motor status data, send the status data of the first drive motor and the second drive motor to the opposite motor controller respectively.

[0145] S702. Compare the speeds of the two motors to calculate the real-time speed difference, analyze the cause of the speed error and calculate the corresponding compensation torque to obtain the first compensation torque command and the second compensation torque command.

[0146] S703. Optimize the motor control process according to the first compensation torque command and the second compensation torque command.

[0147] In this embodiment, based on the dual-motor status data, the status data of the first and second drive motors are sent to the corresponding motor controllers. The left and right motor controllers, as two independent units in the vehicle network, continuously collect real-time status data of their respective motors, including but not limited to the actual rotor speed obtained through a high-precision encoder or rotary transformer, and the real-time output torque estimated by a current sensor and motor model. Each controller periodically encapsulates the collected status data, along with a timestamp generated by its local clock, into a standard data frame. This data frame is transmitted via a dedicated high-speed communication link connecting the two controllers. This link can be implemented based on a deterministic and low-latency vehicle network protocol, such as a dedicated virtual channel for vehicle Ethernet. Upon receiving the data frame, the receiving controller ensures the security and integrity of the data through security checks and aligns the timestamp in the data frame with the time of its own data collection to ensure temporal consistency of the status data on both sides. If no valid data is received from the other side within a preset time window, the receiver will trigger a communication fault flag. The time window can be set according to the control accuracy requirements.

[0148] Specifically, the system compares the speeds of the two motors to calculate the real-time speed difference, analyzes the causes of speed errors, and calculates the corresponding compensation torque, resulting in the first and second compensation torque commands. Through the analysis of the causes of speed errors, appropriate compensation algorithms can be called to differentiate between different deviation situations, avoiding poor performance or oscillations caused by using a single strategy to handle all problems, thus improving the effectiveness and efficiency of compensation. Rapid identification and high dynamic response compensation for sudden tire slippage can intervene immediately when wheels slip, improving response speed, effectively preventing power loss and vehicle dynamic instability, and enhancing vehicle stability and safety under complex road conditions.

[0149] The motor control process is optimized according to the first and second compensation torque commands. By compensating the torque through compensation commands, high-frequency disturbances such as speed differences can be dynamically suppressed, ensuring that the compensation commands will not cause actuator overload or destructive torque mutations. The adjustment range is limited within the safety boundary of the physical system, thereby improving the reliability of the system.

[0150] Furthermore, the real-time speed difference is calculated by comparing the speeds of the two motors, the causes of the speed error are analyzed, and the corresponding compensation torque is calculated to obtain the first compensation torque command and the second compensation torque command, including:

[0151] S801. Compare the speeds of the two motors to calculate the real-time speed difference, analyze the causes of speed error and calculate the corresponding weights. The causes of speed error include load differences, electromechanical characteristic differences and tire slippage.

[0152] S802. Calculate the load compensation torque for load difference, the efficiency compensation torque for electromechanical characteristic difference, and the slip compensation torque for tire slippage, and perform a weighted summation according to the weights to obtain the first compensation torque command and the second compensation torque command.

[0153] In this embodiment, the controller continuously receives the left and right motor speed values ​​exchanged via high-speed communication and time-aligned, calculates their real-time difference and the rate of change of this difference, and combines this with the vehicle's steering wheel angle, yaw rate, current total drive torque command, and historical speed difference data. It analyzes whether the vehicle is in a significant steering state. If the steering wheel angle is small, it mainly considers load differences and electromechanical characteristic differences. It learns and updates a reference deviation table recording the inherent speed difference patterns caused by manufacturing and assembly tolerances under different total torque commands. If the current speed difference matches the predicted value of the reference deviation table, it is considered an electromechanical characteristic difference; if there is a stable offset that matches known vehicle load distribution information, it is considered a load difference; if the vehicle is in a significant steering state, it calculates the theoretical inner and outer wheel speed difference. If the rate of change of the speed difference exceeds a preset dynamic threshold, it is considered tire slippage. The dynamic threshold can be set according to the system control accuracy. Based on the cause strength, a confidence weight between 0 and 1 is calculated for each of the three causes: load difference, electromechanical characteristic difference, and tire slippage. The sum of the weights is 1.

[0154] Specifically, the reference deviation table is a two-dimensional lookup table that takes the total torque command as input and the corresponding inherent speed difference as output. It records the speed difference patterns of the dual-motor system under ideal conditions, excluding external load differences, tire slippage, and other dynamic disturbances, due to inherent factors such as manufacturing tolerances, assembly deviations, and inconsistent motor characteristics. This table is usually calibrated during vehicle production through no-load bench testing or flat road drag testing: under the condition of the vehicle traveling at a constant speed in a straight line, different total torque commands are gradually applied, and the steady-state speed difference between the left and right motors at that torque is recorded, filtered, and stored as a reference value. During online operation, the controller compares the current measured speed difference with the reference value under the corresponding total torque command in the table. If the deviation is within the preset tolerance range, it is determined to be a normal electromechanical characteristic difference.

[0155] It should be noted that by analyzing the causes of speed errors, the root causes of problems can be accurately identified, and compensation for specific causes can improve the effectiveness and efficiency of overall control. It can also analyze the dynamic changes in speed difference caused by different operating conditions such as no load, heavy load, straight driving, turning, and uneven road surfaces, ensuring that the compensation strategy always matches the current state.

[0156] Specifically, the controller operates three compensation calculation channels in parallel. For the load difference compensation channel, a proportional-integral controller (PIC) is used, taking the real-time speed difference as input. The proportional part provides a fast correction response, while the integral part is used to gradually eliminate the steady-state speed deviation caused by continuous asymmetrical load. For the electromechanical characteristic difference compensation channel, a preset two-dimensional lookup table is used for analysis and compensation. The row index of the two-dimensional lookup table is the average motor torque command, the column index is the average motor speed, and the stored value is the torque compensation amount required to offset the characteristic difference of the opposite motor. The corresponding compensation value is obtained by looking up this table based on the current operating point. For the tire slip compensation channel, an algorithm emphasizing rapid suppression is adopted. The algorithm calculates the rate of change of the speed difference in real time. Once the rate exceeds the set sensitivity threshold, a proportional-derivative controller with high gain is activated to generate a compensation torque proportional to the rate of change. The sensitivity threshold can be set according to the system control accuracy requirements.

[0157] Specifically, the pre-defined two-dimensional lookup table construction process includes: In bench tests or towing tests, the left and right motors are mounted on a test bench, mechanically coupled via a coupling or connected to independent dynamometers; under the same speed and torque command, the difference in actual output torque between the left and right motors is measured; the test covers typical operating areas of the vehicle: average motor torque command from minimum to maximum; average motor speed from minimum to maximum; at each operating point, the additional torque correction required to maintain speed synchronization on both sides is recorded, and the sign and magnitude of the correction are the compensation amount. The compensation amounts for all operating points are then filled into a two-dimensional array to obtain the two-dimensional lookup table.

[0158] Preferably, after calculating the three compensation torque values, they are multiplied by their corresponding weights and summed to obtain the compensation torque requirement. In order to ensure that the compensation action mainly generates yaw moment to correct the vehicle's attitude rather than significantly changing the vehicle's longitudinal driving force, the comprehensive compensation torque requirement is transformed into a pair of equal-sized and opposite-direction first compensation torque commands and second compensation torque commands, indicating that while increasing the torque of one motor, the torque of the other motor is reduced by an equal amount.

[0159] It should be noted that by analyzing the causes of the differences and making targeted compensations, the system can make optimal responses to both fast and slow disturbances. This avoids the difficulty in balancing dynamic response and steady-state accuracy when using a single controller parameter. The fast compensation mechanism based on the rate of change, designed for tire slippage, makes the system's response to wheel loss of traction more sensitive and direct than traditional systems based on slip ratio calculations. It can intervene in power distribution earlier, effectively preventing dynamic instability caused by excessive power on one side of the vehicle, enhancing driving safety on wet and loose roads, and improving the vehicle's safety boundary under extreme conditions.

[0160] Furthermore, according to the first compensation torque command and the second compensation torque command, the motor control process is optimized, including:

[0161] S901. The first compensation torque command and the second compensation torque command are respectively superimposed on the first torque command to obtain the second torque command;

[0162] S902. According to the second torque command, the motor control process is adjusted and optimized in real time.

[0163] In this embodiment, the controller simultaneously acquires two sets of inputs: a first torque command, including the basic target torque values ​​for both the left and right motors; and a first compensation torque command and a second compensation torque command, which are typically a pair of torque couples of similar magnitude but opposite directions. The second torque command for the left motor is equal to the sum of its first torque command and the first compensation torque command, and the second torque command for the right motor is also equal to the sum of its first torque command and the second compensation torque command. The calculated left and right second torque commands are compared with the maximum driving torque and maximum braking torque that the motor can safely output at the current actual speed to ensure they meet the safety range. If they exceed the safety range, the second torque commands are adjusted and reduced to obtain the final second torque command.

[0164] It should be noted that superposition can correct minor deviations between actuators, and safety verification can ensure that the instructions issued to the motor power hardware are always within the safety envelope of the physical actuators, thereby improving the safety and stability of mining vehicle operations.

[0165] Specifically, after receiving the second torque command, the motor controller determines the q-axis current reference value required to achieve the target torque, and the d-axis current reference value that may be needed to extend the high-speed operating range, based on the current motor speed and DC bus voltage status, by looking up a table. These current reference values ​​are the tracking targets for the inner current loop. The controller samples the motor's three-phase current in real time and obtains the actual d-axis and q-axis currents in the synchronous rotating coordinate system through coordinate transformation. The current loop controller, typically a proportional-integral regulator, calculates the current tracking error of the d-axis and q-axis and outputs d-axis and q-axis voltage compensation amounts to correct the error through adjustment calculations. Based on the compensated voltage command, inverse transformation and pulse width modulation are performed, and the inverter power switch is controlled to drive the motor to output the desired torque.

[0166] It should be noted that by optimizing the motor control process, it is possible to ensure that the actual torque output is highly consistent with the command, effectively suppress torque fluctuations or oscillations caused by compensation, ensure smooth and stable power output, achieve rapid and stable torque adjustment, and ensure that the dual-motor drive system can maintain a precise, efficient, and stable collaborative working state under various operating conditions.

[0167] like Figure 4 As shown, a dual-motor cooperative control system for a mining dual-driver trackless rubber-tired vehicle is used to implement a dual-motor cooperative control method for a mining dual-driver trackless rubber-tired vehicle, including:

[0168] The status analysis module collects real-time operating status data, driver operation data, and roadway environment data of the dual-driver trackless rubber-tired vehicle for mining, analyzes the real-time status of the vehicle, and constructs the first state vector.

[0169] The control command generation module configures a dual-motor cooperative control mechanism, analyzes the first state vector to construct state constraints, performs coupled analysis on the longitudinal and lateral forces of the vehicle, calculates the target torques of the first drive motor and the second drive motor respectively, and obtains the first torque command, wherein the target torque includes the target drive torque and the target braking torque.

[0170] The dual-motor collaborative control module, in response to the first torque command, controls the working process of the first drive motor and the second drive motor respectively, and collects the motor status data in real time to obtain dual-motor status data.

[0171] The control optimization module sends the status data of the first drive motor and the second drive motor to the corresponding motor controller based on the dual motor status data. It compares the speeds of the two motors and calculates the corresponding compensation torque. The control process of the motor is optimized according to the compensation torque to achieve coordinated control of the two motors.

[0172] Example 3:

[0173] This embodiment illustrates the overall workflow of the technical solution in a specific application scenario. A dual-driver trackless rubber-tired mining vehicle is traveling in a damp underground tunnel, about to enter a curve with a small radius of curvature. At this moment, the driver intends to accelerate.

[0174] Various sensors on the vehicle begin to operate: wheel speed sensors measure the rotational speed of each wheel; the inertial measurement unit (IMU) acquires the vehicle's longitudinal acceleration and yaw rate; the accelerator pedal position sensor detects that the pedal opening is 50%; the steering wheel angle sensor measures the steering wheel angle at 10 degrees; the front-mounted lidar detects that the starting point of the curve is 30 meters ahead, and there are local water stains on the road surface; simultaneously, the UWB-based positioning module, combined with a pre-stored high-precision digital map, confirms that the vehicle's current position is 20 meters from the curve entrance, the curve's radius of curvature is 30 meters, and the road surface's designed slope is 2%. All this multi-source, heterogeneous raw data is sent to the vehicle domain controller. The controller first timestamps each data point and transforms the obstacle position information detected by the lidar into a coordinate system with the vehicle's center of mass as the origin using a pre-calibrated coordinate transformation matrix, completing time synchronization and spatial alignment. The controller analyzes and integrates this information: based on wheel speed and IMU data, it estimates the vehicle's current longitudinal speed to be 20 km / h and the yaw rate to be close to 0; based on a 50% pedal opening, it obtains the driver's expected longitudinal acceleration value of 1.5 m / s² through a lookup table; based on a 10-degree steering wheel angle and vehicle speed, it calculates the driver's expected direct lateral turning angle; simultaneously, based on the 30-meter curve curvature radius provided by the map, it calculates the expected reference lateral turning angle required to smoothly navigate the curve. The controller integrates these key parameters to construct a digitized first state vector, represented as: [Vehicle speed: 20 km / h, expected longitudinal acceleration: 1.5 m / s², steering wheel angle: 10 degrees, reference path curvature: 1 / 30, vehicle-curvature distance to curve entrance: 20 m, road surface adhesion coefficient estimate: 0.35 (wet road surface)], reflecting that the vehicle is approaching a slippery curve at a speed of 20 km / h, the driver wants to accelerate, but the road curvature requires steering.

[0175] The control command generation module receives the first state vector and performs demand calculations: based on the expected longitudinal acceleration of 1.5 m / s², vehicle mass, and gradient, the total longitudinal force required to achieve this acceleration is calculated to be 8000 Newtons. Simultaneously, to coordinate the driver's steering input with the curve curvature requirements and maintain vehicle stability, an additional yaw moment of 500 Newton-meters is calculated. Safety boundary calculations are then performed: based on the estimated wet road adhesion coefficient of 0.35 and real-time vehicle load, the maximum longitudinal force available for driving each drive tire without sideslip is calculated. Simultaneously, the motor external characteristic performance spectrum is consulted to obtain the maximum torque that each motor can provide at the current speed. Combining these two constraints, a dynamic feasible region is plotted on the total driving force-yaw moment two-dimensional plane. Under the current wet road surface and vehicle speed, the maximum allowable total driving force in this feasible region is 10000 Newtons, and the maximum yaw moment is 800 Newton-meters. The target point, defined by a longitudinal demand force of 8000 N·m and a yaw moment of 500 N·m, falls precisely within the feasible region, indicating that the driver's acceleration and steering intentions can be safely achieved under the current physical conditions. By solving the dynamic equations, this target point is precisely decomposed into torque commands for the left and right motors: the calculated drive torque required by the left motor is 400 N·m, and the drive torque required by the right motor is 550 N·m. These two torque values ​​are then encapsulated into the first torque command.

[0176] The dual-motor coordinated control module receives the first torque command and drives the left and right motor controllers to control their respective motors to output the target torque. Simultaneously, the control optimization module begins operation. The left and right motor controllers exchange their status data in real time via a high-speed communication link, focusing on the actual motor speeds. Due to a wetter road surface on the right or a slight difference in the characteristics of the right motor, after outputting the torque command, the actual speed of the right motor was detected to be 501 rpm, while the left motor's was 498 rpm, a difference of 3 rpm. The control optimization module immediately analyzes this speed difference: considering the vehicle's slight steering and the slippery road surface, it determines that the difference is mainly caused by tire slippage. A compensation torque is calculated: a +5 Nm compensation command is generated for the left motor, and a -5 Nm compensation command is generated for the right motor. These two compensation commands are immediately superimposed on the original first torque command. The final command for the left motor is adjusted to 405 Nm, and for the right motor, it is adjusted to 545 Nm. This adjustment is equivalent to fine-tuning the power distribution between the left and right wheels while keeping the total driving force basically unchanged, generating a small yaw moment that suppresses the tendency to slip on the right side.

[0177] As the vehicle enters and accelerates through a slippery curve, it responds to the driver's acceleration intention and actively ensures the stability of the curve. Starting with the analysis of the environment and the driver's intention, it calculates the optimal torque distribution within safety constraints through coupled dynamics analysis, and overcomes real-time disturbances through actuator closed-loop compensation within milliseconds. This achieves safe, stable, efficient and driver-expected passability, improving the intelligent control level and operational safety of mining vehicles in complex underground environments.

[0178] The above description is merely a preferred embodiment of this application. The scope of protection of this application is not limited to the above embodiments. All technical solutions falling within the scope of this application's concept are within the scope of protection of this application. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of this application should also be considered within the scope of protection of this application.

Claims

1. A dual-motor coordinated control method for a dual-driver trackless rubber-tired vehicle used in mining, characterized in that, include: Real-time data collection of operating status, driver operation data, and roadway environment data of dual-driver trackless rubber-tired vehicles for mining; analysis of real-time vehicle status; and construction of the first state vector. A dual-motor cooperative control mechanism is configured, state constraints are constructed by analyzing the first state vector, and the longitudinal and lateral forces of the vehicle are coupled and analyzed. The target torques of the first drive motor and the second drive motor are calculated respectively to obtain the first torque command. The target torque includes the target drive torque and the target braking torque. In response to the first torque command, the operation of the first drive motor and the second drive motor are controlled respectively, and the status data of the motors are collected in real time to obtain the status data of the two motors. Based on the dual-motor status data, the status data of the first drive motor and the second drive motor are sent to the motor controller on the other side, the speeds of the two motors are compared and the corresponding compensation torque is calculated, and the motor control process is optimized according to the compensation torque in order to perform coordinated control of the dual motors.

2. The dual-motor coordinated control method for a dual-driver trackless rubber-tired vehicle for mining, as described in claim 1, is characterized in that... The system collects real-time operating status data, driver operation data, and roadway environment data of the dual-driver trackless rubber-tired mining vehicle, analyzes the vehicle's real-time status, and constructs a first state vector, including: Real-time acquisition of operating status data, driver operation data, and roadway environment data of dual-driver trackless rubber-tired vehicles for mining, resulting in multi-source vehicle data; Multi-source vehicle data is synchronized in time and aligned in space. The geometric parameters of the roadway, vehicle pose parameters and environmental obstacle parameters are analyzed to construct the first state vector.

3. The dual-motor coordinated control method for a dual-driver trackless rubber-tired mining vehicle according to claim 1, characterized in that, The dual-motor cooperative control mechanism includes: The first state vector is analyzed to construct state constraints. The longitudinal and lateral dynamics of the vehicle are coupled and analyzed to calculate the longitudinal demand force and yaw moment of the vehicle. The dynamic feasible region of the dual motor joint operation is calculated based on the state constraints. Target points are constructed based on longitudinal demand force and yaw moment. The target torques of the first and second drive motors are calculated based on the dynamic feasible domain to obtain the target torque command set. Based on the target torque command set, the braking and driving requirements of the corresponding motor are decomposed, the target driving torque or target braking torque is calculated, and the first torque command is obtained.

4. The dual-motor coordinated control method for a dual-driver trackless rubber-tired mining vehicle according to claim 3, characterized in that, The analysis of the first state vector constructs state constraints, performs coupled analysis of the vehicle's longitudinal and lateral dynamics, calculates the vehicle's longitudinal demand force and yaw moment, and calculates the dynamic feasible region for the joint operation of the two motors based on the state constraints, including: Based on the analysis of the first state vector, the expected longitudinal acceleration is obtained, and the expected lateral turning angle is obtained by analyzing the tunnel curvature information. Based on the real-time vehicle status and roadway conditions, calculate the longitudinal demand force and yaw moment corresponding to the expected longitudinal acceleration and expected lateral turning angle, respectively; Calculate the maximum lateral force that prevents the vehicle tires from slipping laterally, and combine the maximum lateral force with the maximum longitudinal force of the two motors to construct state constraints; Based on the state constraints, the dynamic feasible region for the joint operation of the two motors is calculated.

5. The dual-motor coordinated control method for a dual-driver trackless rubber-tired mining vehicle according to claim 3, characterized in that, The target point is constructed according to the longitudinal demand force and the yaw moment. The target torque of the first drive motor and the second drive motor is calculated based on the dynamic feasible region, resulting in a target torque command set, including: Construct the target point based on the longitudinal demand force and the yaw moment; Within the dynamic feasible domain, the target torque of the first drive motor and the second drive motor are calculated according to the position of the target point to obtain a set of target torque commands. If the target point is not within the dynamic feasible region, the target point is projected onto the dynamic feasible region to obtain the projection point. The target torque of the first drive motor and the second drive motor are calculated according to the position of the projection point to obtain the target torque command set.

6. The dual-motor coordinated control method for a dual-driver trackless rubber-tired mining vehicle according to claim 3, characterized in that, The step of decomposing the braking and driving requirements of the corresponding motor according to the target torque command set, calculating the target driving torque or target braking torque, and obtaining the first torque command includes: The target torque with a negative value is selected from the target torque command set, the braking demand is analyzed, and the braking demand is decomposed into the first braking part and the second braking part according to the real-time status of the motor. Based on the real-time status of the motors, the first braking component is dynamically allocated to each motor, the corresponding target braking torque is calculated, and the second braking component is allocated to the vehicle's hydraulic control component. Based on the positive target torque, analyze the drive demand and calculate the target drive torque of the motor; The first torque command is obtained by combining the target driving torque or the target braking torque.

7. The dual-motor coordinated control method for a dual-driver trackless rubber-tired vehicle for mining, as described in claim 1, is characterized in that... The process of sending the status data of the first and second drive motors to the corresponding motor controller based on the dual-motor status data, comparing the speeds of the two motors and calculating the corresponding compensation torque, and optimizing the motor control process according to the compensation torque includes: Based on the dual motor status data, the status data of the first drive motor and the second drive motor are sent to the motor controller on the other side respectively; The speed difference between the two motors is calculated by comparing their speeds, the cause of the speed error is analyzed and the corresponding compensation torque is calculated, and the first compensation torque command and the second compensation torque command are obtained. The motor control process is optimized according to the first compensation torque command and the second compensation torque command.

8. The dual-motor coordinated control method for a dual-driver trackless rubber-tired vehicle for mining, as described in claim 7, is characterized in that... The process involves comparing the speeds of the two motors, calculating the real-time speed difference, analyzing the causes of the speed error, calculating the corresponding compensation torque, and obtaining a first compensation torque command and a second compensation torque command, including: The speed difference between the two motors is calculated by comparing their speeds in real time. The causes of the speed error are analyzed and the corresponding weights are calculated. The causes of the speed error include load differences, electromechanical characteristic differences and tire slippage. The load compensation torque for load difference, the efficiency compensation torque for electromechanical characteristic difference, and the slip compensation torque for tire slip are calculated, and then weighted and summed according to their respective weights to obtain the first compensation torque command and the second compensation torque command.

9. The dual-motor coordinated control method for a dual-driver trackless rubber-tired vehicle for mining, as described in claim 7, is characterized in that... The process of optimizing the motor control according to the first compensation torque command and the second compensation torque command includes: The first compensation torque command and the second compensation torque command are respectively superimposed on the first torque command to obtain the second torque command; The motor control process is adjusted and optimized in real time according to the second torque command.

10. A dual-motor coordinated control system for a dual-driver trackless rubber-tired mining vehicle, characterized in that, A dual-motor coordinated control method for implementing a dual-driver trackless rubber-tired vehicle for mining as described in any one of claims 1 to 9 includes: The status analysis module collects real-time operating status data, driver operation data, and roadway environment data of the dual-driver trackless rubber-tired vehicle for mining, analyzes the real-time status of the vehicle, and constructs the first state vector. The control command generation module configures a dual-motor cooperative control mechanism, analyzes the first state vector to construct state constraints, performs coupled analysis on the longitudinal and lateral forces of the vehicle, calculates the target torques of the first drive motor and the second drive motor respectively, and obtains the first torque command, wherein the target torque includes the target drive torque and the target braking torque. The dual-motor collaborative control module, in response to the first torque command, controls the working process of the first drive motor and the second drive motor respectively, and collects the motor status data in real time to obtain dual-motor status data. The control optimization module sends the status data of the first drive motor and the second drive motor to the corresponding motor controller based on the dual motor status data. It compares the speeds of the two motors and calculates the corresponding compensation torque. The control process of the motor is optimized according to the compensation torque to achieve coordinated control of the two motors.

Images