A multi-motor driving torque coordination control method of an electric forklift

Abstract

This invention relates to the field of electric construction machinery drive control technology, specifically a multi-motor drive torque coordination control method for electric shovels; it includes data acquisition, model building, parameter mapping, and torque adjustment processes; the system constructs a virtual viscoelastic shaft model based on the Ackermann principle; its core is to map real-time steering angle, load, and speed difference into virtual stiffness and damping coefficients, calculate the coupling compensation torque, and superimpose it onto the base torque; this invention breaks the traditional opposition between electronic differential and anti-slip control, and achieves flexible connection between the left and right motors through the characteristics of the virtual shaft, effectively suppressing slippage without cutting off power, and taking into account both steering flexibility and driving stability under complex working conditions.

Description

Technical Field

[0001] This invention relates to the field of electric construction machinery drive control technology, specifically a multi-motor drive torque coordination control method for electric loaders. Background Technology

[0002] In the current heavy-duty operation environment of electric forklifts, the vehicles often need to perform high-frequency turning and straight-line driving under complex road conditions such as wet, slippery and soft surfaces, and the multi-motor drive system needs to respond to changes in load and road conditions in real time.

[0003] To ensure vehicle trajectory and power output, existing solutions generally employ electronic differentials based on geometric models or anti-slip control technologies based on slip ratio thresholds. While these solutions can maintain basic operation in specific scenarios, electronic differentials and anti-slip control have inherently conflicting control objectives. Traditional electronic differentials lack physical connection constraints, which can easily cause severe wheel spin and power loss on low-traction surfaces. Conventional anti-slip strategies often rely on abrupt power cut-offs or forced speed synchronization, resulting in steering sluggishness and reduced vehicle traction. They are unable to balance steering agility and driving stability, and cannot effectively address the risk of center of gravity instability caused by drastic load changes.

[0004] Therefore, how to eliminate the control deadlock between differential smoothness and anti-slip safety, and achieve torque adaptive coordination and stable output under multiple operating conditions, has become an urgent technical problem to be solved. Summary of the Invention

[0005] To solve the above-mentioned technical problems, the present invention provides a multi-motor drive torque coordination control method for electric loaders. Specifically, the technical solution of the present invention includes:

[0006] Step 1: Collect real-time speed data of the left drive motor, right drive motor, steering angle of the steering mechanism, and load mass data of the lifting system during the operation of the electric loader. Based on the Ackermann steering geometry principle, construct the theoretical speed difference between the left and right drive wheels and establish a virtual viscoelastic shaft model that includes virtual stiffness coefficient and virtual damping coefficient.

[0007] Step 2: Construct stiffness mapping function and damping mapping function. Input the real-time steering angle and real-time load mass into the stiffness mapping function to generate the virtual stiffness coefficient under the current working condition. At the same time, calculate the rate of change of the difference between the real-time speed of the left drive motor and the real-time speed of the right drive motor, and input it into the damping mapping function to generate the virtual damping coefficient under the current working condition.

[0008] Step 3: Based on the deviation between the theoretical speed difference and the actual speed difference, and combined with the virtual stiffness coefficient and the virtual damping coefficient, calculate the virtual internal torque generated by the virtual viscoelastic shaft model to obtain the coupling compensation torque. Then, add the coupling compensation torque to the basic torque requested by the driver to generate the target torque command for the left motor and the target torque command for the right motor. In response to the target torque commands for the left motor and the right motor, adjust the current vector of the inverter to drive the electric loader.

[0009] Preferably, step one includes:

[0010] S11. Using a rotary transformer or encoder installed on the motor shaft, obtain the real-time speed of the left drive motor and the real-time speed of the right drive motor respectively, and calculate the actual speed difference between the two.

[0011] S12. Use an angle sensor to collect the real-time steering angle of the steering mechanism, and use an oil pressure sensor to collect the pressure data of the lifting system and convert it into real-time load mass.

[0012] S13. Based on the wheelbase and track parameters of the electric loader, combined with the real-time steering angle, calculate the ideal kinematic relationship of the vehicle in a slip-free state, construct the theoretical speed difference, and map the left drive motor and the right drive motor into an equivalent physical model connected by a virtual viscoelastic axis, and establish a virtual viscoelastic axis model.

[0013] Preferably, step two includes:

[0014] S21. Extract the absolute value of the real-time steering angle and construct a stiffness attenuation factor that is monotonically decreasing with respect to the absolute value of the steering angle.

[0015] S22. Extract the real-time load mass and construct a stiffness gain factor that is monotonically increasing with the load mass;

[0016] S23. Multiply or sum the stiffness attenuation factor and stiffness gain factor to generate a virtual stiffness coefficient through the stiffness mapping function. In the straight driving state where the real-time steering angle approaches zero, the virtual stiffness coefficient tends to infinity.

[0017] S24. Differentiate the actual speed difference to obtain the rate of change of speed difference, which serves as a slip characteristic quantity characterizing the wheel slipping trend.

[0018] S25. Input the slip characteristic quantity into the damping mapping function to generate a virtual damping coefficient, wherein the virtual damping coefficient has a monotonically increasing nonlinear relationship with the amplitude of the slip characteristic quantity.

[0019] Preferably, step three includes:

[0020] S31. Calculate the speed deviation between the actual speed difference and the theoretical speed difference;

[0021] S32. Integrate the rotational speed deviation to obtain the torsional deformation of the virtual viscoelastic shaft;

[0022] S33. Multiply the virtual stiffness coefficient by the torsional deformation to obtain the elastic recovery component, and multiply the virtual damping coefficient by the rotational speed deviation to obtain the viscous hysteresis component.

[0023] S34. Add the elastic recovery component and the viscous resistance component to obtain the coupling compensation torque, wherein the direction of the coupling compensation torque is defined as the direction that hinders the expansion of the speed deviation.

[0024] Preferably, step three also includes:

[0025] S35. Obtain the total drive torque requested by the driver via the accelerator pedal and distribute it evenly to the base torque;

[0026] S36. Subtract the coupling compensation torque from the base torque of the left motor to generate the target torque command for the left motor;

[0027] S37. Add the coupling compensation torque to the base torque of the right motor to generate the target torque command for the right motor;

[0028] S38. Convert the target torque command of the left motor and the target torque command of the right motor into dq axis current commands respectively, and use space vector pulse width modulation technology to control the on and off of the inverter power switching transistors.

[0029] Preferably, in step S23, the virtual stiffness coefficient is evaluated to obtain a first working condition evaluation result, wherein the preset limit steering threshold is numerically greater than the preset straight-line threshold, and the evaluation includes:

[0030] When the absolute value of the real-time steering angle is less than or equal to the preset straight-line threshold, it is determined to be a straight-line driving condition. The virtual stiffness coefficient is set to the preset maximum stiffness value, and the left and right drive motor speeds are forced to synchronize.

[0031] When the absolute value of the real-time steering angle is greater than the preset straight line threshold and less than or equal to the preset limit steering threshold, it is determined to be a normal steering condition. The virtual stiffness coefficient is dynamically adjusted according to the stiffness mapping function, allowing the left and right drive motors to generate a speed difference that conforms to Ackermann geometry.

[0032] When the absolute value of the real-time steering angle is greater than the preset limit steering threshold, it is determined to be a stationary steering condition. The virtual stiffness coefficient is set to the preset minimum stiffness value, and the speed coupling between the left and right drive motors is decoupled.

[0033] Preferably, in step S25, the slip characteristic quantity is evaluated to obtain a second working condition evaluation result, wherein the preset slip threshold is numerically greater than the preset micro-slip threshold, and the evaluation includes:

[0034] When the slip characteristic is less than or equal to the preset micro-slip threshold, it is determined to be a stable adhesion condition, and the virtual damping coefficient is set to the preset basic damping value to suppress low-frequency oscillations.

[0035] When the slip characteristic is greater than the preset micro-slip threshold and less than or equal to the preset slip threshold, it is determined to be a potential slip condition, and the virtual damping coefficient is controlled to increase linearly with the slip characteristic.

[0036] When the slip characteristic quantity is greater than the preset slip threshold, it is judged as a severe slip condition. The virtual damping coefficient is controlled to increase exponentially to generate a high-intensity viscous resistance component to suppress the slipping wheel from spinning.

[0037] Preferably, the method further includes an energy redistribution strategy for severe slippage conditions:

[0038] S41. When the condition is determined to be a severe slippage condition, extract the high-intensity viscous damping component generated by the virtual damping coefficient.

[0039] S42. Identify the high-speed slipping side motor and the low-speed non-slipping side motor;

[0040] S43. By coupling compensation torque, part of the torque of the slipping side motor is transferred to the non-slipping side motor, and the adhesion of the non-slipping side motor is used to maintain the vehicle driving force, thereby realizing the electronic limited-slip differential function.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1. This method breaks the traditional opposition between electronic differential and anti-slip control. By introducing a physically isomorphic virtual viscoelastic shaft model, the left and right drive motors are regarded as connected by a virtual shaft with variable torsional stiffness and variable viscous damping. The virtual stiffness characteristic is used to ensure the vehicle's trajectory tracking ability under different operating conditions and simulate the synchronization of mechanical connection. The virtual damping characteristic utilizes the natural dissipation principle of the physical model to respond to the wheel slippage tendency within milliseconds and suppress wheel spin by generating viscous drag components. This design does not require the abrupt cut-off of power as in traditional anti-slip strategies, thus enabling the vehicle to maintain both steering flexibility and driving stability under complex wet and slippery conditions.

[0043] 2. To address the unique large-angle operation requirements of electric forklifts, this method establishes a stiffness adaptive adjustment mechanism based on real-time steering angle. During straight-line travel, the left and right motor speeds are forced to synchronize by setting a maximum stiffness value, eliminating overshoot oscillations and improving tracking performance. During normal steering, the stiffness is dynamically adjusted according to Ackermann geometry, allowing natural differential speed. In particular, for stationary steering, the virtual stiffness is automatically reduced to a minimum to decouple the vehicle. This segmented control avoids the motor overheating or jamming problems caused by calculating singularities in traditional electronic differentials at large angles, significantly improving the vehicle's maneuverability in confined spaces.

[0044] 3. This method abandons the single threshold control and instead uses a graded approach based on the magnitude of the speed difference change rate. In the micro-slip stage, linear damping is used to suppress low-frequency oscillations, and in the severe slip stage, exponentially increasing damping force is used to quickly curb idling. More importantly, this method can execute an energy redistribution strategy in the case of severe slip, automatically transferring the power that was wasted by the slipping motor to the non-slipping motor through coupling compensation torque. This electronic limited-slip differential function ensures the maintenance of the vehicle's total traction under extreme road conditions, effectively solving the risk of getting stuck caused by power cut-off in traditional anti-slip systems, and significantly improving the vehicle's ability to get out of trouble and its off-road performance.

[0045] 4. Considering the characteristics of electric loaders, such as drastic load changes and a high center of gravity during operation, this method introduces the real-time load mass data of the lifting system into the control closed loop and constructs a stiffness gain factor that is positively correlated with the load. As the load increases, the system automatically increases the stiffness coefficient of the virtual shaft, thereby enhancing the synchronous coupling strength of the left and right drive wheels. This design enables the control system to sense and adapt to changes in the vehicle's center of gravity, effectively preventing the risk of instability or rollover caused by insufficient stiffness when the high center of gravity vehicle is turning under heavy load or driving on complex road surfaces, and providing proactive safety assurance. Attached Figure Description

[0046] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0047] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0049] Example 1:

[0050] Please see Figure 1 A method for coordinated control of multi-motor drive torque in an electric loader, comprising the following steps:

[0051] Step 1: Collect real-time speed data of the left drive motor, right drive motor, steering angle of the steering mechanism, and load mass data of the lifting system during the operation of the electric loader. Based on the Ackermann steering geometry principle, construct the theoretical speed difference between the left and right drive wheels and establish a virtual viscoelastic shaft model that includes virtual stiffness coefficient and virtual damping coefficient.

[0052] Step 2: Construct stiffness mapping function and damping mapping function. Input the real-time steering angle and real-time load mass into the stiffness mapping function to generate the virtual stiffness coefficient under the current working condition. At the same time, calculate the rate of change of the difference between the real-time speed of the left drive motor and the real-time speed of the right drive motor, and input it into the damping mapping function to generate the virtual damping coefficient under the current working condition.

[0053] Step 3: Based on the deviation between the theoretical speed difference and the actual speed difference, and combined with the virtual stiffness coefficient and the virtual damping coefficient, calculate the virtual internal torque generated by the virtual viscoelastic shaft model to obtain the coupling compensation torque. Then, add the coupling compensation torque to the basic torque requested by the driver to generate the target torque command for the left motor and the target torque command for the right motor. In response to the target torque commands for the left motor and the right motor, adjust the current vector of the inverter to drive the electric loader.

[0054] This embodiment provides a multi-motor drive torque coordination control method for electric loader. The core of this method is to break the traditional opposition between electronic differential and anti-skid control. By introducing a physically isomorphic virtual viscoelastic shaft model, adaptive adjustment of vehicle dynamic characteristics is achieved. The system executes a data acquisition step to obtain key state variables through a sensor network. The source is a sensor at the motor shaft end, and its physical meaning is the real-time speed of the left and right drive motors, with the unit being rad / s; The source is a steering angle sensor, and its physical meaning is the real-time steering angle of the steering mechanism, measured in rad. The source is hydraulic circuit pressure conversion, and its physical meaning is the real-time load mass of the lifting system, in kg. The system constructs a virtual viscoelastic shaft model at the algorithm level. This model treats the left and right drive motors as connected through a virtual viscoelastic shaft with variable torsional stiffness and variable viscous damping, aiming to simulate the physical characteristics of mechanical connections. Based on this, the system calls stiffness mapping functions and damping mapping functions to generate virtual stiffness coefficients that determine the synchronization strength. And the virtual damping coefficient that determines the disturbance rejection capability Based on the deviation between the theoretical speed difference and the actual speed difference, the coupling compensation torque caused by the deformation of the virtual shaft is calculated. The compensated torque is superimposed on the base torque to generate the final motor torque command, which drives the inverter to adjust the current vector.

[0055] In this embodiment, under heavy-duty operation scenarios of electric forklifts, a physical model including virtual stiffness and damping is established, which effectively solves the deadlock problem in the prior art where differential smoothness and anti-slip safety cannot be simultaneously achieved. Virtual stiffness ensures the vehicle's trajectory tracking ability under different loads and steering conditions, while virtual damping responds to slippage trends within milliseconds. The natural damping characteristics of the physical model suppress wheel spinning without abruptly cutting off power, thereby enabling the vehicle to operate smoothly under complex and slippery conditions.

[0056] Example 2:

[0057] Step one includes:

[0058] S11. Using a rotary transformer or encoder installed on the motor shaft, obtain the real-time speed of the left drive motor and the real-time speed of the right drive motor respectively, and calculate the actual speed difference between the two.

[0059] S12. Use an angle sensor to collect the real-time steering angle of the steering mechanism, and use an oil pressure sensor to collect the pressure data of the lifting system and convert it into real-time load mass.

[0060] S13. Based on the wheelbase and track parameters of the electric loader, combined with the real-time steering angle, calculate the ideal kinematic relationship of the vehicle in a slip-free state, construct the theoretical speed difference, and map the left drive motor and the right drive motor into an equivalent physical model connected by a virtual viscoelastic axis, and establish a virtual viscoelastic axis model.

[0061] This embodiment further specifies the data acquisition and model building steps in Embodiment 1, clarifying the acquisition path of multi-source heterogeneous data and the foundation for building the theoretical model; the system utilizes a high-precision rotary transformer installed on the motor shaft end to acquire the real-time speed of the left drive motor with a sampling period of microseconds. Real-time speed of the right drive motor The difference between the two values ​​is then calculated. To suppress the influence of high-frequency noise from the sensor on subsequent differential calculations, a second-order low-pass Butterworth filter is used to smooth the speed signal after calculating the actual speed difference. The cutoff frequency is set between 50Hz and 100Hz to ensure the slip characteristic value. The computational stability; The source is calculated by subtracting rotational speeds, and its physical meaning is the actual difference in rotational speed, measured in rad / s. Simultaneously, the system utilizes an absolute encoder to acquire the real-time steering angle of the steering mechanism. The system uses an oil pressure sensor installed in the lifting hydraulic cylinder circuit to collect pressure data, which is then converted into real-time load mass through a calibration curve. Based on the key geometric parameters of electric forklifts—wheelbase This refers to the distance between the center planes of the left and right wheels, which in this embodiment ranges from 1.8m to 2.2m, with a typical value of 2.0m, and is related to the wheelbase. This refers to the distance between the centerlines of the front and rear axles. In this embodiment, the value ranges from 2.8m to 3.2m, with a typical value of 3.0m. Combined with the real-time steering angle, the ideal kinematic relationship of the vehicle in a slip-free state is calculated to construct the theoretical speed difference. The specific calculation formula is as follows:

[0062]

[0063] The negative sign is introduced in the formula to adapt to the definition of the steering angle sign, setting left turn as the positive direction. At this point, the theoretical rotational speed of the left wheel is... It should be less than the speed of the right wheel. That is, the theoretical speed difference should be negative; it should be noted that this formula for calculating the theoretical speed difference is derived based on ideal Ackermann steering geometry and is applicable to the linear or quasi-linear region of the vehicle's steering angle, for example... When the steering angle is close to Under extreme operating conditions, a nonlinear correction factor needs to be introduced, but within the normal operating range of this embodiment, the linear approximation meets the control accuracy requirements.

[0064] in, and These represent the real-time speeds of the left and right drive motors, respectively, with subscripts... Represents Left. Represents Right; Item This formula represents the vehicle's current center-of-motion average speed, derived under small-angle approximation conditions based on Ackermann steering geometry; B represents the track width, and L represents the wheelbase. The source is calculated using the above formula, and its physical meaning is the theoretical speed difference, with the unit being rad / s. The system maps the left and right motors as equivalent physical models connected by virtual viscoelastic shafts, providing a reference system for subsequent deviation calculations.

[0065] In this embodiment, under the dynamic control scenario of an electric loader, by collecting rotational speed at high frequency and combining it with the Ackerman geometry construction theory reference, it is possible to accurately distinguish between normal steering differential speed and abnormal slip differential speed. In particular, the introduction of load mass collection enables the control system to sense changes in the vehicle's center of gravity, which plays a crucial safety role in preventing the high center of gravity electric loader from tipping over when steering under heavy load.

[0066] Example 3:

[0067] Step two includes:

[0068] S21. Extract the absolute value of the real-time steering angle and construct a stiffness attenuation factor that is monotonically decreasing with respect to the absolute value of the steering angle.

[0069] S22. Extract the real-time load mass and construct a stiffness gain factor that is monotonically increasing with the load mass;

[0070] S23. Multiply or sum the stiffness attenuation factor and stiffness gain factor to generate a virtual stiffness coefficient through the stiffness mapping function. In the straight driving state where the real-time steering angle approaches zero, the virtual stiffness coefficient tends to infinity.

[0071] S24. Differentiate the actual speed difference to obtain the rate of change of speed difference, which serves as a slip characteristic quantity characterizing the wheel slipping trend.

[0072] S25. Input the slip characteristic quantity into the damping mapping function to generate a virtual damping coefficient, wherein the virtual damping coefficient has a monotonically increasing nonlinear relationship with the amplitude of the slip characteristic quantity.

[0073] This embodiment further specifies the parameter generation logic in Embodiment 2, and elaborates on how to endow the virtual axis with adaptive adjustment capability; the system extracts the absolute value of the real-time steering angle and constructs a stiffness attenuation factor. This factor decreases as the steering angle increases, and its specific functional form is:

[0074]

[0075] in, Steering attenuation coefficient, unit: Its physical meaning is to characterize the rate at which virtual stiffness decreases exponentially with increasing steering angle, and its value range is... Determined based on vehicle steering sensitivity testing;

[0076] The calibration method for the key coefficient is disclosed here: attenuation coefficient. Used to adjust the rate of stiffness reduction during steering, with a value range set to [value range]. In this embodiment, it is preferred that This value ensures that the stiffness decays to approximately 20% of the base value at the maximum steering angle, balancing differential requirements and self-centering torque; simultaneously, real-time load mass is extracted to construct a stiffness gain factor. This factor increases with increasing load, and its specific functional form is:

[0077]

[0078] in, The unit mass stiffness gain coefficient, in units of: Used to linearly characterize load quality Enhancement effect on the system's basic stiffness requirements; gain coefficient Based on the ratio of the vehicle's fully loaded to its unloaded mass, it is set to... This unit is used to offset Dimensions of mass, product term It is a dimensionless number used to match the dimensions of the constant term; for example, when the load mass When the load is 2500kg, the system stiffness increases by 1.5 times; by combining the two through a stiffness mapping function, a virtual stiffness coefficient is generated. The specific calculation formula is as follows:

[0079]

[0080] in, Based on the stiffness, As the turning attenuation factor, This is the load gain coefficient; The source is a mapping function calculation, and its physical meaning is a virtual stiffness coefficient, with units of Nm / rad; among which, the basic stiffness... Set as Nm / rad, this value is based on the motor's rated torque With the maximum permissible angle of twist ratio This is determined; during this process, in response to the real-time steering angle approaching zero, the system sets the virtual stiffness coefficient to approach infinity to simulate the differential lock effect; the system performs differential processing on the actual speed difference to obtain the speed difference change rate, specifically using a first-order backward difference algorithm for discretization calculation:

[0081]

[0082] in, For the current moment The calculated instantaneous slip characteristic quantities, referred to as variables in this paper, are... ; The subscript represents the actual difference in rotational speed between the left and right wheels detected at the current moment. Represents Actual; The previous sampling time The actual speed difference, subscript Represents Previous; The sampling time interval is set to [value]. To match the motor current loop update frequency;

[0083] The sampling time interval is disclosed here. The specific value is s, or 5ms, is a frequency that matches the current loop update cycle of the motor controller to ensure no signal aliasing. The source is the aforementioned differential calculation, and its physical meaning is the slip characteristic quantity, with units of rad / s². This slip characteristic quantity is input into the damping mapping function to generate the virtual damping coefficient. As a basic implementation of continuous control, this coefficient is designed to have a monotonically increasing nonlinear relationship with the amplitude of the slip characteristic, specifically in the form of:

[0084]

[0085] The above-mentioned index model is the basic implementation method of damping adjustment. In more complex operating conditions, in order to take into account the smoothness under different degrees of slippage, a segmented mapping strategy as described in the example can be used for fine adjustment.

[0086] in, Basic damping, Sensitivity coefficient; designed for rapid response to sudden slippage; basic damping. Set as Nms / rad, used to eliminate steady-state micro-oscillations; sensitivity coefficient Set as This ensures that the damping force dominates the torque output when the slip characteristic reaches 20 rad / s².

[0087] In this embodiment, a nonlinear parameter space is constructed for complex road conditions. The stiffness coefficient is designed to balance the tracking performance of straight-line driving with the flexibility of steering, and the stability is enhanced by adaptive load. The exponential response design of the damping coefficient makes the system extremely sensitive to even the slightest signs of slippage, and can generate huge viscous resistance in the early stage of slippage, thus achieving proactive safety control to prevent slippage before it occurs.

[0088] Example 4:

[0089] Step three includes:

[0090] S31. Calculate the speed deviation between the actual speed difference and the theoretical speed difference;

[0091] S32. Integrate the rotational speed deviation to obtain the torsional deformation of the virtual viscoelastic shaft;

[0092] S33. Multiply the virtual stiffness coefficient by the torsional deformation to obtain the elastic recovery component, and multiply the virtual damping coefficient by the rotational speed deviation to obtain the viscous hysteresis component.

[0093] S34. Add the elastic recovery component and the viscous resistance component to obtain the coupling compensation torque, wherein the direction of the coupling compensation torque is defined as the direction that hinders the expansion of the speed deviation.

[0094] This embodiment further specifies the torque calculation algorithm in Embodiment 3, illustrating how to calculate physical torque using virtual parameters; the system calculates the difference between the actual speed difference and the theoretical speed difference to obtain the speed deviation. Integrating this deviation yields the torsional deformation of the virtual viscoelastic axis. The integration process employs the following discretization formula:

[0095]

[0096] in, Torsion deformation, Time index Sampling period;

[0097] Set the initial conditions for integration as follows: The source is the aforementioned integral calculation, and its physical meaning is the torsional deformation of the virtual viscoelastic shaft, in rad. The system calculates the elastic recovery component and the viscous hysteresis component separately: the virtual stiffness coefficient is multiplied by the torsional deformation to obtain the former, and the virtual damping coefficient is multiplied by the rotational speed deviation to obtain the latter; these two components are added together to synthesize the coupled compensation torque. ;

[0098]

[0099] in, This represents the speed deviation. Its integral over time;

[0100] This formula uses a Kelvin-Voigt-like model, where, The simulation of the elastic torsional restoring force of the shaft system, Simulated damping dissipation force; The source is the summation of components, and its physical meaning is the coupling compensation torque, with the unit being Nm. The direction of this torque is strictly defined as the direction that hinders the expansion of the speed deviation, so as to form a negative feedback control closed loop.

[0101] To ensure long-term system stability and prevent numerical drift caused by the integral component, this embodiment also includes integral reset logic: when the system detects the steering angle... The actual speed difference between the left and right wheels is zero and continues. When the value falls below a preset silent threshold, the currently accumulated torsional deformation will be automatically adjusted. Reset to zero;

[0102] In this embodiment, a variant of the Kelvin-Voigt model based on physical meaning is used in the torque synthesis process. Compared with pure mathematical PID control, this torque synthesis method based on a physical model has a clear energy meaning. The elastic term provides steady-state accuracy to ensure steering geometry, and the viscous term provides transient stability to suppress slippage oscillation. This combination effectively eliminates the overshoot oscillation phenomenon commonly found in traditional control and improves the control quality.

[0103] Example 5:

[0104] Step three also includes:

[0105] S35. Obtain the total drive torque requested by the driver via the accelerator pedal and distribute it evenly to the base torque;

[0106] S36. Subtract the coupling compensation torque from the base torque of the left motor to generate the target torque command for the left motor;

[0107] S37. Add the coupling compensation torque to the base torque of the right motor to generate the target torque command for the right motor;

[0108] S38. Convert the target torque command of the left motor and the target torque command of the right motor into dq axis current commands respectively, and use space vector pulse width modulation technology to control the on and off of the inverter power switching transistors.

[0109] This embodiment is a further specification of the torque distribution and execution strategy in Embodiment 4; the system analyzes the signal input by the driver through the accelerator pedal to obtain the requested total drive torque. and distribute it evenly to the base torque. ; The source is calculated by equal distribution, and its physical meaning is the basic torque of a single-sided motor, with units of Nm; the system performs torque reconstruction: subtracting the coupling compensation torque from the basic torque of the left motor to generate the target torque command for the left motor. Add coupling compensation torque to the base torque of the right motor to generate the target torque command for the right motor. In this step, the system applies a non-negative saturation limit or regenerative braking limit to the generated target torque command, i.e., when the calculated... or When the value is less than zero, its amplitude is limited to a preset reverse braking safety threshold range, for example... Or clamp directly to This is to prevent vehicle instability caused by wheel reversal due to overcompensation; among which, the symbol and These represent the final target torque command values ​​for the left and right motors, respectively, indicated by the superscript. This represents the target setpoint of the control system, and the controller will drive the motor to operate according to this instruction.

[0110] These two target torque commands are input to the field-oriented control module of the motor controller, converted into dq-axis current commands, and the inverter power switching transistors are controlled to turn on and off using space vector pulse width modulation technology, thereby generating driving force at the physical level.

[0111] In a multi-motor drive scenario, this embodiment essentially implements an electronic limited-slip differential function. By superimposing complementary torques of opposite values ​​on the base torque, the system achieves dynamic redistribution of torque between the left and right wheels without changing the total driving force, ensuring accurate execution of the driver's acceleration intentions, while avoiding power interruption caused by unilateral braking intervention, thus improving the vehicle's power responsiveness.

[0112] Example 6:

[0113] In step S23, the virtual stiffness coefficient is evaluated to obtain the evaluation result for the first working condition. The preset limit steering threshold is numerically greater than the preset straight-line threshold. The evaluation includes:

[0114] When the absolute value of the real-time steering angle is less than or equal to the preset straight-line threshold, it is determined to be a straight-line driving condition. The virtual stiffness coefficient is set to the preset maximum stiffness value, and the left and right drive motor speeds are forced to synchronize.

[0115] When the absolute value of the real-time steering angle is greater than the preset straight line threshold and less than or equal to the preset limit steering threshold, it is determined to be a normal steering condition. The virtual stiffness coefficient is dynamically adjusted according to the stiffness mapping function, allowing the left and right drive motors to generate a speed difference that conforms to Ackermann geometry.

[0116] When the absolute value of the real-time steering angle is greater than the preset limit steering threshold, it is determined to be a stationary steering condition. The virtual stiffness coefficient is set to the preset minimum stiffness value, and the speed coupling between the left and right drive motors is decoupled.

[0117] This embodiment further specifies the virtual stiffness coefficient control strategy in Embodiment 3, refining the segmented control logic for different steering conditions. The system monitors the absolute value of the steering angle in real time and compares it with a preset threshold. In response to a steering angle less than the straight-line threshold, the system determines it to be a straight-line driving condition. At this time, in order to eliminate the torque abrupt change caused by directly switching from the maximum stiffness value to the mapped stiffness value, the system introduces a transition weighting function to smooth the virtual stiffness coefficient. The specific calculation formula is as follows:

[0118]

[0119] in, This refers to the final output virtual stiffness coefficient. The original stiffness is calculated based on the formula in Example 3; The transition weight function is defined as follows:

[0120]

[0121] in, This is the extreme turning threshold. The threshold is a straight line; this weighting logic ensures that when... Just over When the weights are close to 1, the stiffness is smoothly maintained at... Nearby; as the steering angle increases, the weight decreases linearly to 0, and the stiffness transitions smoothly to... This ensures the dynamic stability of the vehicle when entering a curve;

[0122] At this point, the virtual stiffness coefficient is set to the maximum stiffness value to establish a rigid connection and force the left and right drive motors to synchronize their speeds. When the steering angle is between the straight-line threshold and the extreme steering threshold, the system determines it as a normal steering condition. In this case, the coefficient is dynamically adjusted according to the stiffness mapping function to allow for a speed difference conforming to Ackermann geometry. When the steering angle exceeds the extreme steering threshold, the system determines it as a stationary steering condition and sets the virtual stiffness coefficient to the minimum stiffness value, thereby decoupling the speed. The numerical definition of the key stiffness extreme value is disclosed here: the preset maximum stiffness value. Set as Nm / rad, this value simulates the rigid connection characteristics of a mechanical differential lock when locked; preset minimum stiffness value Set as Nm / rad, which means completely disconnecting the virtual shaft connection and eliminating internal stress;

[0123] This embodiment addresses the unique large-angle operation requirements of electric forklifts by employing a segmented stiffness control strategy to resolve the conflict between straight-line stability and maneuverability in stationary turning. In particular, the decoupling design for stationary turning avoids the motor overheating or jamming issues caused by singularity calculations in traditional electronic differentials at large angles, significantly improving the vehicle's maneuverability in confined spaces. To enable those skilled in the art to implement this method appropriately, the specific setting logic and values ​​of key thresholds are disclosed here: Straight-line threshold. It is based on the measurement of mechanical clearance in the steering hydraulic system, and is set as follows: rad, this value can effectively filter signal noise caused by unintentional minor adjustments to the steering wheel by the driver; extreme steering threshold It is determined based on the linear boundary of the vehicle's Ackerman steering geometry, and is set as follows: rad; Through vehicle dynamics simulation verification, compared with the fixed stiffness strategy, the adaptive control strategy using the above threshold reduced the standard deviation of the yaw rate of the vehicle when traveling in a straight line at 40km / h by 72%, effectively eliminating overshoot oscillation.

[0124] Example 7:

[0125] In step S25, the slip characteristic quantity is evaluated to obtain the evaluation result of the second working condition, wherein the preset slip threshold is numerically greater than the preset micro-slip threshold, and the evaluation includes:

[0126] When the slip characteristic is less than or equal to the preset micro-slip threshold, it is determined to be a stable adhesion condition, and the virtual damping coefficient is set to the preset basic damping value to suppress low-frequency oscillations.

[0127] When the slip characteristic is greater than the preset micro-slip threshold and less than or equal to the preset slip threshold, it is determined to be a potential slip condition, and the virtual damping coefficient is controlled to increase linearly with the slip characteristic.

[0128] When the slip characteristic quantity is greater than the preset slip threshold, it is judged as a severe slip condition. The virtual damping coefficient is controlled to increase exponentially to generate a high-intensity viscous resistance component to suppress the slipping wheel from spinning.

[0129] To ensure the continuity of the damping force at the slippage threshold and avoid torque step shocks, a modified exponential model based on the offset is used here. The specific calculation formula is as follows:

[0130]

[0131] in, To limit the damping reference value, used to ensure minimum system damping; The damping gain coefficient determines the amplitude of the damped response; its unit is set to [unit missing]. , with virtual damping coefficient Maintain consistency of dimensions; It is an exponential growth factor used to adjust the damping effect on slip deviation. The sensitivity level is set in units of [missing information]. This is to ensure that the exponent term is a dimensionless value; This specifically refers to the set slippage detection threshold; This represents the current virtual damping coefficient. For the current slip characteristic, this formula ensures that when hour, This achieves a smooth transition from the linear growth region to the exponential growth region;

[0132] To enable those skilled in the art to implement this control strategy, the key parameter setting in this embodiment is disclosed herein: damping gain coefficient. Set as Exponential growth factor Set as This parameter combination ensures that the slip characteristic quantity Exceeding the slip threshold After Within the virtual damping coefficient, a peak damping force sufficient to suppress idling can be quickly established;

[0133] This embodiment further specifies the virtual damping coefficient control strategy in Embodiment 3, refining the hierarchical management logic for different degrees of slippage. Unlike the single continuous exponential model used in Embodiment 3, this embodiment employs a more refined piecewise control strategy to optimize ride comfort under different operating conditions. The system evaluates slip characteristics in real time. The magnitude of the characteristic quantity; in response to the characteristic quantity being less than the micro-slip threshold, the system determines a stable adhesion condition and sets the virtual damping coefficient to the base damping value to suppress the low-frequency mechanical resonance of the motor itself; in response to the characteristic quantity being between the micro-slip threshold and the slip threshold, the system determines a potential slip condition and controls the virtual damping coefficient to increase linearly with the slip characteristic quantity to provide moderate damping, the function expression of which is:

[0134]

[0135] in, Limiting damping, Slip threshold Microslip threshold; limiting damping Values Nms / rad ensures the continuity of damping force at the slip threshold; in response to the characteristic quantity exceeding the slip threshold, the system determines that it is a severe slip condition and controls the virtual damping coefficient to increase exponentially to generate a high-intensity viscous drag component.

[0136] This embodiment employs a layered management strategy in anti-slip control scenarios. It intervenes gently during slight slippage, without affecting the driving experience; and intervenes decisively during severe slippage, rapidly dissipating the kinetic energy of the slipping wheel using exponentially increasing damping force. Compared to the trigger-and-cut-off control of traditional TCS, this solution provides a continuous and smooth anti-slip experience, ensuring the vehicle's controllability in critical adhesion states. The specific threshold settings are based on the following: micro-slip threshold. Set as rad / s², this value is determined based on the measured value of the maximum speed fluctuation rate of a vehicle under full load during rapid acceleration on a dry asphalt road surface; slippage threshold. Set as rad / s², this value corresponds to the rotational acceleration characteristic when the tire slip ratio exceeds the 0.2 peak point; experimental data shows that at low adhesion coefficients In road tests, this parameter setting resulted in a system response delay of less than 20ms, successfully limiting wheel slippage speed to within 105% of the target value.

[0137] Example 8:

[0138] The method also includes energy redistribution strategies for severe slippage conditions:

[0139] S41. When the condition is determined to be a severe slippage condition, extract the high-intensity viscous damping component generated by the virtual damping coefficient.

[0140] S42. Identify the high-speed slipping side motor and the low-speed non-slipping side motor;

[0141] S43. By coupling compensation torque, part of the torque of the slipping side motor is transferred to the non-slipping side motor, and the adhesion of the non-slipping side motor is used to maintain the vehicle driving force, thereby realizing the electronic limited-slip differential function.

[0142] This embodiment further specifies the severe slippage handling strategy in Embodiment 7, describing the specific execution process of energy redistribution. After the system determines that a severe slippage condition has been entered, the high-intensity viscous drag component generated by the virtual damping coefficient is extracted. The system identifies the slipping motor with a higher speed and the non-slipping motor with a lower speed by comparing the speeds of the left and right motors. Utilizing the physical characteristics of coupled compensation torque, an energy transfer operation is performed: the torque output of the slipping motor is significantly reduced to suppress idling, while the torque output of the non-slipping motor is increased by an equal amount. The effective adhesion of the non-slipping motor is used to maintain the vehicle's total driving force.

[0143] This embodiment achieves true electronic limited-slip differential functionality in extreme off-road or stuck-in-the-ground scenarios. Unlike traditional mechanical differentials that lose power when slipping, this solution automatically transfers the wasted power from the slipping motor to the non-slipping motor using a virtual viscoelastic shaft model. This not only suppresses slippage but, more importantly, maintains the vehicle's total traction under harsh road conditions, significantly improving the electric loader's ability to get out of trouble and its off-road performance. To verify the effectiveness of this technology, a single-wheel ice surface traction comparison test was conducted: the left wheel was placed on the ice surface. The right wheel was placed on the cement road surface. On a 15% slope, a vehicle with conventional electronic differential control experiences severe left wheel spin and is unable to climb the slope; however, the vehicle using the method described in this embodiment has its left wheel speed clamped below 30 rad / s, and the right wheel receives an additional 40% torque compensation. The vehicle successfully starts and completes the climb within 1.5 seconds, proving that this method solves the deadlock problem and significantly improves the vehicle's ability to get out of trouble.

[0144] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for coordinated control of multi-motor drive torque in an electric loader, characterized in that, The specific steps include: Step 1: Collect real-time speed data of the left drive motor, right drive motor, steering angle of the steering mechanism, and load mass data of the lifting system during the operation of the electric loader. Based on the Ackermann steering geometry principle, construct the theoretical speed difference between the left and right drive wheels and establish a virtual viscoelastic shaft model that includes virtual stiffness coefficient and virtual damping coefficient. Step 2: Construct stiffness mapping function and damping mapping function. Input the real-time steering angle and real-time load mass into the stiffness mapping function to generate the virtual stiffness coefficient under the current working condition. At the same time, calculate the rate of change of the difference between the real-time speed of the left drive motor and the real-time speed of the right drive motor, and input it into the damping mapping function to generate the virtual damping coefficient under the current working condition. Step 3: Based on the deviation between the theoretical speed difference and the actual speed difference, and combined with the virtual stiffness coefficient and the virtual damping coefficient, calculate the virtual internal torque generated by the virtual viscoelastic shaft model to obtain the coupling compensation torque. Then, add the coupling compensation torque to the basic torque requested by the driver to generate the target torque command for the left motor and the target torque command for the right motor. In response to the target torque commands for the left motor and the right motor, adjust the current vector of the inverter to drive the electric loader. Step three includes: S31. Calculate the speed deviation between the actual speed difference and the theoretical speed difference; S32. Integrate the rotational speed deviation to obtain the torsional deformation of the virtual viscoelastic shaft; S33. Multiply the virtual stiffness coefficient by the torsional deformation to obtain the elastic recovery component, and multiply the virtual damping coefficient by the rotational speed deviation to obtain the viscous hysteresis component. S34. Add the elastic recovery component and the viscous resistance component to obtain the coupling compensation torque, wherein the direction of the coupling compensation torque is defined as the direction that hinders the expansion of the speed deviation.

2. The multi-motor drive torque coordination control method for an electric loader according to claim 1, characterized in that, Step one includes: S11. Using a rotary transformer or encoder installed on the motor shaft, obtain the real-time speed of the left drive motor and the real-time speed of the right drive motor respectively, and calculate the actual speed difference between the two. S12. Use an angle sensor to collect the real-time steering angle of the steering mechanism, and use an oil pressure sensor to collect the pressure data of the lifting system and convert it into real-time load mass. S13. Based on the wheelbase and track parameters of the electric loader, combined with the real-time steering angle, calculate the ideal kinematic relationship of the vehicle in a slip-free state, construct the theoretical speed difference, and map the left drive motor and the right drive motor into an equivalent physical model connected by a virtual viscoelastic axis, and establish a virtual viscoelastic axis model.

3. The multi-motor drive torque coordination control method for an electric loader according to claim 2, characterized in that, Step two includes: S21. Extract the absolute value of the real-time steering angle and construct a stiffness attenuation factor that is monotonically decreasing with respect to the absolute value of the steering angle. S22. Extract the real-time load mass and construct a stiffness gain factor that is monotonically increasing with the load mass; S23. Multiply or sum the stiffness attenuation factor and stiffness gain factor to generate a virtual stiffness coefficient through the stiffness mapping function. In the straight driving state where the real-time steering angle approaches zero, the virtual stiffness coefficient tends to infinity. S24. Differentiate the actual speed difference to obtain the rate of change of speed difference, which serves as a slip characteristic quantity characterizing the wheel slipping trend. S25. Input the slip characteristic quantity into the damping mapping function to generate a virtual damping coefficient, wherein the virtual damping coefficient has a monotonically increasing nonlinear relationship with the amplitude of the slip characteristic quantity.

4. The multi-motor drive torque coordination control method for an electric loader according to claim 1, characterized in that, Step three also includes: S35. Obtain the total drive torque requested by the driver via the accelerator pedal and distribute it evenly to the base torque; S36. Subtract the coupling compensation torque from the base torque of the left motor to generate the target torque command for the left motor; S37. Add the coupling compensation torque to the base torque of the right motor to generate the target torque command for the right motor; S38. Convert the target torque command of the left motor and the target torque command of the right motor into dq axis current commands respectively, and use space vector pulse width modulation technology to control the on and off of the inverter power switching transistors.

5. The multi-motor drive torque coordination control method for an electric loader according to claim 3, characterized in that, In step S23, the virtual stiffness coefficient is evaluated to obtain the evaluation result for the first working condition, wherein the preset limit steering threshold is numerically greater than the preset straight-line threshold. The evaluation includes: When the absolute value of the real-time steering angle is less than or equal to the preset straight-line threshold, it is determined to be a straight-line driving condition. The virtual stiffness coefficient is set to the preset maximum stiffness value, and the left and right drive motor speeds are forced to synchronize. When the absolute value of the real-time steering angle is greater than the preset straight line threshold and less than or equal to the preset limit steering threshold, it is determined to be a normal steering condition. The virtual stiffness coefficient is dynamically adjusted according to the stiffness mapping function, allowing the left and right drive motors to generate a speed difference that conforms to Ackermann geometry. When the absolute value of the real-time steering angle is greater than the preset limit steering threshold, it is determined to be a stationary steering condition. The virtual stiffness coefficient is set to the preset minimum stiffness value, and the speed coupling between the left and right drive motors is decoupled.

6. The multi-motor drive torque coordination control method for an electric loader according to claim 3, characterized in that, In step S25, the slip characteristic quantity is evaluated to obtain the second working condition evaluation result, wherein the preset slip threshold is numerically greater than the preset micro-slip threshold, and the evaluation includes: When the slip characteristic is less than or equal to the preset micro-slip threshold, it is determined to be a stable adhesion condition, and the virtual damping coefficient is set to the preset basic damping value to suppress low-frequency oscillations. When the slip characteristic is greater than the preset micro-slip threshold and less than or equal to the preset slip threshold, it is determined to be a potential slip condition, and the virtual damping coefficient is controlled to increase linearly with the slip characteristic. When the slip characteristic quantity is greater than the preset slip threshold, it is judged as a severe slip condition. The virtual damping coefficient is controlled to increase exponentially to generate a high-intensity viscous resistance component to suppress the slipping wheel from spinning.

7. A multi-motor drive torque coordination control method for an electric loader according to claim 6, characterized in that, The method also includes an energy redistribution strategy for severe slippage conditions: S41. When the condition is determined to be a severe slippage condition, extract the high-intensity viscous damping component generated by the virtual damping coefficient. S42. Identify the high-speed slipping side motor and the low-speed non-slipping side motor; S43. By coupling compensation torque, part of the torque of the slipping side motor is transferred to the non-slipping side motor, and the adhesion of the non-slipping side motor is used to maintain the vehicle driving force, thereby realizing the electronic limited-slip differential function.

Images