A method and system for motor torque zero-crossing control

Abstract

This application relates to a method and system for controlling the zero-crossing torque of a motor. It generates a smooth target torque curve segment corresponding to the complete zero-crossing process, constructs a displacement power function, and adjusts parameters t, a, and n to make a preset curve segment in the curve formed by the displacement power function approximate the smooth target torque curve segment, thus obtaining the target curve. Upon receiving a zero-crossing command, the target curve is retrieved and the real-time torque value is output. First, t independently controls the horizontal shift of the torque response curve, ensuring that the torque remains at zero or near zero within a buffer window of settable duration after zero-crossing, providing a shock-free clearance for gear meshing and resolving the contradiction between smoothness and response speed caused by the inability to independently control delays in existing technologies. Second, the joint adjustment of a and n makes the power function curve segment approximate a smoother target curve segment than the original demand curve, truly reflecting the torque transmission characteristics from gear disengagement to re-engagement, eliminating gear impact caused by sudden torque changes at a physical level, and avoiding gear grinding noise and vibration.

Description

Technical Field

[0001] This application relates to the field of new energy transportation technology, and in particular to a method and system for controlling the zero-crossing torque of a motor. Background Technology

[0002] With the increasing popularity of electric vehicles, the control precision and driving comfort of motor drive systems are receiving growing attention. In the transmission chain of an electric vehicle, the torque output by the motor is transmitted to the drive wheels via a reduction gear set. Due to the inherent backlash in mechanical components such as gears and splines, when the motor torque demand changes from a positive value across zero to a negative value, such as when releasing the accelerator or during regenerative braking, or when the motor torque demand changes from a negative value across zero to a positive value, such as during start-up or acceleration, instantaneous impacts and collisions occur between the gear pairs, a phenomenon known as "gear grinding." Gear grinding not only generates significant noise and vibration, severely reducing ride comfort, but may also accelerate fatigue wear of the transmission system.

[0003] To suppress gear grinding impact when the torque crosses zero, traditional solutions employ quadratic functions. Fitting the zero-crossing curve or using an exponential function y=aˣ to control the torque change rate are both methods. However, these approaches have several inherent drawbacks. First, quadratic functions require simultaneous adjustment of both parameters a and b, necessitating segmentation (from negative to 0, and from 0 to positive), necessitating at least two sets of a and b parameters, making vehicle tuning very complex. Second, sometimes the vehicle requires a zero-crossing torque delay from the MCU, which quadratic power functions cannot achieve, failing to provide targeted waiting for the actual time window of gear grinding, making it difficult to balance smoothness and response speed. Furthermore, traditional solutions do not consider the motor's own torque response capability, such as the maximum torque change rate that the motor controller can actually output. The generated ideal curve may exceed the motor's physical limits, causing control commands to be inaccurately executed and potentially causing secondary impacts.

[0004] Therefore, there is an urgent need for a motor torque zero-crossing control method that can flexibly adjust the torque change pattern in the zero-crossing range, independently control the response delay time, and actively adapt to the motor response capability, so as to eliminate gear impact while taking into account driving responsiveness and calibration efficiency. Summary of the Invention

[0005] Therefore, it is necessary to provide a method and system for controlling the zero-crossing of motor torque to address the aforementioned problems of traditional solutions.

[0006] On the one hand, this application provides a method for controlling the zero-crossing of motor torque, including: Generate a custom curve; the custom curve includes a smooth target torque curve segment, which corresponds in time to the complete process of the target output torque changing from the first polarity across zero to the second polarity; Construct a power function of displacement , where y is the fitted output torque, x is time, t is the delay response time parameter, a is the rate of change parameter, n is the trend of change parameter, and the curve corresponding to the displacement power function is defined as the displacement power function curve; By adjusting the delay response time parameter, the rate of change parameter, and the trend of change parameter, a preset curve segment of the displacement power function curve is made to approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve. When a control command is received that the motor's required torque changes from the first polarity across zero to the second polarity, the target displacement power function curve is retrieved, and the real-time torque value is output based on the target displacement power function curve.

[0007] This application relates to a method for controlling the zero-crossing of motor torque, which involves generating a smooth target torque curve segment that corresponds in time to the complete process of the target output torque changing from a first polarity across zero to a second polarity, and then constructing a displacement power function. The curve is defined as a displacement power function curve. By adjusting the delay response time parameter t, the rate of change parameter a, and the trend parameter n, a preset segment of the displacement power function curve is made to approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve. Finally, when a zero-crossing command is received, the target displacement power function curve is retrieved and the real-time torque value is output. The solution of this application achieves the following technical effects: First, by using the independent parameter t to achieve horizontal translation of the torque response curve, the torque is kept at zero or near zero for a settable time after zero crossing, providing a shock-free buffer window for gear meshing, solving the defect of traditional methods that cannot independently control the delay time, resulting in difficulty in balancing smoothness and response speed; Second, by adjusting the parameters a and n, the displacement power function curve segment is made to approximate the smooth target torque curve segment in the custom curve. This smooth target torque curve segment has a slower initial slope change compared to the original motor torque demand curve, more realistically reflecting the torque transmission characteristics during the gear disengagement and re-engagement process, thereby eliminating gear impact caused by sudden torque changes at the physical level and avoiding gear knocking noise and vibration caused by directly following the demand curve. Thus, without increasing hardware costs, a smooth transition to zero torque was achieved, and the adaptability of the control strategy to different vehicle models was significantly improved.

[0008] On the other hand, this application also provides a motor torque zero-crossing control system, including: Electric motor; A reduction gear set is mechanically connected to the output shaft of the motor; A torque sensor is used to detect the output torque or torque demand of the motor in real time. One or more controllers, the controllers being communicatively connected to the torque sensor and configured to execute the motor torque zero-crossing control method mentioned above; When the controller receives a control command that changes the required torque of the motor from the first polarity to the second polarity across zero, it retrieves the target displacement power function curve and outputs a real-time torque value based on the target displacement power function curve to reduce impact during the torque zero-crossing engagement of the reduction gear set.

[0009] This application relates to a motor torque zero-crossing control system. By constructing a closed-loop control system including a motor, a reduction gear set, a torque sensor, and a controller, and configuring the controller to execute the motor torque zero-crossing control method mentioned above, a complete hardware and software collaboration is achieved, from custom curve generation and power function approximation to real-time torque output. This control system, without adding additional mechanical structures or sensors, utilizes the existing motor controller's computing power and significantly improves gear meshing impact through software-level algorithm upgrades. It has extremely high engineering practicality and promotional value, and ensures stable and reliable reduction of zero-crossing gear breakage during actual vehicle operation. Attached Figure Description

[0010] Figure 1 This is a flowchart of a method for controlling the zero-crossing torque of a motor, provided in an embodiment of this application.

[0011] Figure 2 This is a schematic diagram of a motor torque zero-crossing control system provided in an embodiment of this application.

[0012] Figure 3 This is a schematic diagram illustrating the gear tooth-gripping principle according to an embodiment of this application.

[0013] Figure 4 This is a schematic diagram of a custom curve in a motor torque zero-crossing control method provided in an embodiment of this application.

[0014] Figure 5 This is a schematic diagram of the target displacement power function curve in a motor torque zero-crossing control method provided in an embodiment of this application.

[0015] Figure 6 This is a schematic diagram comparing the motor torque demand curve, the user-defined curve, and the target displacement power function curve in a motor torque zero-crossing control method provided in an embodiment of this application.

[0016] Figure 7 This is a schematic diagram of the vehicle requested torque curve, the motor response torque curve, and the motor speed curve in a motor torque zero-crossing control method provided in an embodiment of this application.

[0017] Figure label: 10 - Motor; 110 - Motor output shaft; 20 - Reduction gear set; 30 - Torque sensor; 40 - Controller. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] This application provides a method for controlling the zero-crossing of motor torque. It should be noted that this application provides a method for controlling the zero-crossing of motor torque.

[0020] Furthermore, the motor torque zero-crossing control method provided in this application does not limit the executing entity. Optionally, the executing entity of the motor torque zero-crossing control method provided in this application can be a motor torque zero-crossing control system. For the sake of brevity, the motor torque zero-crossing control system is simply referred to as the "system".

[0021] like Figure 1 As shown, in one embodiment of this application, the motor torque zero-crossing control method includes: S100, Generate a custom curve. The custom curve includes a smooth target torque curve segment, which corresponds in time to the complete process of the target output torque changing from the first polarity across zero to the second polarity.

[0022] S200, construct a power function of displacement. Where y is the fitted output torque, x is time, t is the delay response time parameter, a is the rate of change parameter, and n is the trend parameter, and the curve corresponding to the displacement power function is defined as the displacement power function curve.

[0023] S300, by adjusting the delay response time parameter, the rate of change parameter, and the trend of change parameter, a preset curve segment of the displacement power function curve is made to approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve.

[0024] S400: When a control command is received that the motor's required torque changes from the first polarity across zero to the second polarity, the target displacement power function curve is retrieved, and a real-time torque value is output based on the target displacement power function curve.

[0025] Specifically, in step S100, the controller first generates a custom curve in its internal memory. Figure 4 The curve in the text is the custom curve L2. This custom curve contains a key segment, namely the smooth target torque curve segment, in... Figure 4In the diagram, the smooth target torque curve segment, which is the curve segment between points A and B, will be referred to as segment AB below. The time range of this smooth target torque curve segment covers the entire transition process of the target output torque changing from zero in one direction to the opposite direction; this process is called the zero-crossing interval.

[0026] The custom curve is our ideal torque response curve. Ideally, by setting the zero-crossing range, that is, setting a smooth target torque curve segment, we can reduce the impact and collision between gears and thus reduce gear grinding.

[0027] pass Figure 3 This will explain why a smooth target torque curve segment can reduce gear impact. Figure 3 In the scenario, assuming there are two gears, K1 and K2, and the gear motion is as follows: gears B1 and A2 first mesh, then gears B1 and A2 disengage, and finally gears B2 and A2 mesh. In the custom curve L2, the portion with the x-coordinate less than the x-coordinate of point A (the curve segment to the left of point A) corresponds to the meshing state of B1 and A2. Conversely, the portion with the x-coordinate greater than the x-coordinate of point B (the curve segment to the right of point B) corresponds to the meshing state of B2 and A2.

[0028] As we can understand, segment AB represents the gradual disengagement of teeth B1 and A2, followed by the gradual engagement of teeth B2 and A2. Segment AB represents a region with significant torque fluctuations. We aim for a slower torque change within segment AB, which translates to a less rapid increase or decrease in the rate of torque change within this segment. In other words, we want the slope of the curve to change less rapidly. Therefore, we created a custom curve with a zero-crossing interval, where the slope changes slowly from 0. For example... Figure 4 As shown, Figure 4 It is the process of torque changing from negative to positive. In the zero-crossing range, such as the part that has just crossed zero, the slope of the point slowly increases from 0, which can reduce the impact and collision between gears.

[0029] In other words, at the exact moment when the torque is zero, the rate of change of torque over time is zero, which is reflected on the curve as a slope of 0 at the point where the torque is 0. As time progresses to both sides, the absolute value of the rate of change gradually increases. This shape can simulate the natural force process of a gear disengaging and re-engaging, avoiding abrupt changes.

[0030] In step S200, the controller constructs a displacement power function, the mathematical expression of which is: In this function, the independent variable *x* represents time, and the dependent variable *y* represents the fitted output torque. The parameter *t* is the delay response time parameter, used to control the horizontal position of the curve on the time axis. The parameter *a* is the rate of change parameter, used to control the vertical scaling of the curve. The parameter *n* is the trend parameter, used to control the curvature of the curve. The controller names the curve corresponding to this function as a displacement power function curve.

[0031] In step S300, the controller adjusts three parameters, t, a, and n, to make the shape of a certain continuous sub-interval on the displacement power function curve highly similar to the shape of the smooth target torque curve segment generated in step S100. This process is called approximation. As can be seen from the figure, our work aims to... Figure 5 CD segment and Figure 4 The shape of segment AB is consistent. This continuous subinterval is referred to as the preset curve segment in this application, corresponding to the zero-crossing interval. After approximation, this displacement power function curve is determined as the target displacement power function curve.

[0032] Figure 5 This is an example of a target displacement power function curve where a equals 1, t equals 7, and n equals 3. In this case, the target displacement power function curve L3 is a cubic parabola with a ordinate of 0 at x=t and a slope of zero. Furthermore, the curve is symmetrical about the point (t, 0). It can be seen that at this point... Figure 5 Target displacement power function curve and Figure 4 The custom curves are highly similar in shape, making them ideal for approximating smooth target torque curve segments from negative to positive or from positive to negative. More specifically, we are... Figure 5 To approximate the CD segment in the middle Figure 4 In the AB segment, we do not care whether the non-zero-crossing intervals are similar.

[0033] In step S300, adjusting n allows us to anchor the basic shape of the curve. Adjusting a changes the steepness of the curve. Adjusting t shifts the x-coordinate of the curve's zero point to the right, thus achieving a delayed response.

[0034] In step S400, when the vehicle controller sends a command to the motor controller requiring torque to cross zero, such as switching from positive torque to negative torque, the motor controller immediately retrieves the calibrated target displacement power function curve from its memory, and finds the corresponding fitted output torque y on the curve based on the current time. This y value is then sent to the motor as a real-time torque command for execution. For example, if the target displacement power function curve is... At 0.06 seconds after receiving the command, the controller calculates y = 0.000002 N·m, and the motor outputs a very small torque. As time increases, the torque gradually increases, achieving a smooth transition.

[0035] It should be noted that the target output torque in S100, the fitted output torque in S200, and the real-time output torque value in S400 all refer to the same physical quantity, torque, i.e., they all refer to the vertical axis of the curve. However, the target output torque is the vertical axis of L2, the fitted output torque is the vertical axis of L3, and the real-time output torque value refers to the torque output by the motor when outputting torque based on L3.

[0036] In this embodiment, a smooth target torque curve segment is generated that corresponds in time to the complete process of the target output torque changing from the first polarity across zero to the second polarity, and then a displacement power function is constructed. The curve is defined as a displacement power function curve. By adjusting the delay response time parameter t, the rate of change parameter a, and the trend parameter n, a preset segment of the displacement power function curve is made to approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve. Finally, when a zero-crossing command is received, the target displacement power function curve is retrieved and the real-time torque value is output. The solution of this application achieves the following technical effects: First, by using the independent parameter t to achieve horizontal translation of the torque response curve, the torque is kept at zero or near zero for a settable time after zero crossing, providing a shock-free buffer window for gear meshing, solving the defect of traditional methods that cannot independently control the delay time, resulting in difficulty in balancing smoothness and response speed. Second, by adjusting the parameters a and n, the displacement power function curve segment is made to approximate the smooth target torque curve segment in the custom curve. This smooth target torque curve segment has a slower initial slope change compared to the original motor torque demand curve, more realistically reflecting the torque transmission characteristics during the gear disengagement and re-engagement process, thereby eliminating gear impact caused by sudden torque changes at the physical level and avoiding gear grinding noise and vibration caused by directly following the demand curve. Thus, without increasing hardware costs, a smooth transition to zero torque was achieved, and the adaptability of the control strategy to different vehicle models was significantly improved.

[0037] In one embodiment of this application, S300 includes, namely, adjusting the delay response time parameter, the rate of change parameter, and the trend of change parameter to make a preset curve segment of the displacement power function curve approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve, including: S310, t is calibrated as the time of delayed response after the torque crosses zero, so that the real-time torque value is output according to the target displacement power function curve only after waiting for time t after crossing zero, and t is greater than 0.

[0038] S320 calculates n based on the motor's torque response slope, and n is used to control the rate of change of the target output torque to be achieved after the preset curve segment ends.

[0039] Specifically, in step S310, the controller calibrates parameter t. t represents the waiting time from the moment the torque actually crosses zero until the controller begins outputting torque according to the target displacement power function curve. For example, if real-vehicle testing reveals that gear grinding occurs between 30ms and 50ms after the torque crosses zero, then t can be calibrated to 50ms. Within the first 50ms after the zero crossing, the controller outputs torque at 0, or it may not actually be 0, but it will maintain the minimum value of the previous polarity. Only after 50ms does it begin to gradually increase the torque according to the target displacement power function curve. Parameter t must be greater than 0 to ensure a buffer period exists.

[0040] like Figure 6 As shown, L1 is the torque demand curve, with its zero point at (20ms, 0). L2 is a custom curve, also the ideal torque response curve, with its zero point at (40ms, 0). This can be understood as generating a torque demand after issuing an operation command indicating that the torque is about to cross zero. However, normal, ideal torque feedback is not instantaneous; it requires a 20ms delay to provide feedback. Based on the ideal custom curve, we further delay by a time 't', where 't' starts from... Figure 6 It's 30ms. Curve L2 is... Curve L3 is To make the matching relationship between the curves and formulas easier to understand, the unit of x is set to 10 ms, meaning x=1 represents 10 ms and x=2 represents 20 ms. The unit of y is set to 10 N·m, meaning x=1 represents 10 N·m and x=2 represents 20 N·m. Our actual torque response curve uses L3 instead of L2 because after the torque demand arises at zero crossing, L3 delays the torque increase by time t compared to the ideal curve L2. This minimizes the risk of gear grinding.

[0041] In step S320, the controller calculates parameter n based on the motor's own torque response slope. The motor's torque response slope refers to the maximum rate of torque change that the motor controller can achieve under given voltage, temperature, and other conditions. The value of n directly affects the rate of torque change after the target displacement power function curve ends at the preset curve segment. Specifically, when n is large, the curve becomes steeper in the later stages of zero crossing, and the torque increases faster. When n is small, the curve is relatively flat in the later stages of zero crossing. By relating n to the motor's actual response capability, it can be ensured that the rate of torque change after zero crossing does not exceed the motor's physical limits, avoiding command distortion.

[0042] In this embodiment, by defining t as the duration of the delayed response after the torque crosses zero and setting t>0, the torque is output according to the target displacement power function curve only after waiting for t time after the zero crossing, thus clarifying the quantitative relationship between the delay parameter and the response start time. Simultaneously, n is calculated based on the motor's torque response slope and used to control the target output torque change rate required after the preset curve segment ends. This scheme allows the delay duration to be adjusted independently of the curve shape and ensures that the response slope after the zero crossing matches the motor's actual output capacity, avoiding response distortion or control overshoot caused by demand exceeding the motor's capacity.

[0043] In one embodiment of this application, S320 includes calculating n based on the torque response slope of the motor, where n is used to control the rate of change of the target output torque to be achieved after the preset curve segment ends, including: S321, obtain the motor's torque response slope.

[0044] S322, the target slope is defined as the slope that is reached at a preset time after the zero-crossing time node in the target displacement power function curve.

[0045] S323, obtain the initial calibration values ​​of t and a.

[0046] S324, using the torque response slope of the motor as the target slope, the initial calibration value of t as the value of t, and the initial calibration value of a as the value of a, n is determined by solving the nonlinear equation. The nonlinear equation is shown in Formula 1.

[0047] Formula 1.

[0048] Where n is the trend parameter, a is the rate of change parameter, m is the target slope, and t is the response time parameter. The preset time after the zero-crossing time node in the target displacement power function curve.

[0049] Specifically, this embodiment explains how n is calculated.

[0050] In step S321, the controller reads the motor's torque response slope from the motor parameter table or obtains it through self-learning. This torque response slope is typically provided by the motor and controller supplier; for example, the maximum torque change rate of a certain model of motor at 5000 rpm and a bus voltage of 350V is 600 N·m / s.

[0051] In step S322, the controller defines a target slope m, which represents the slope value that the curve needs to reach after a preset time interval from the zero-crossing time node (x=t) in the target displacement power function curve. For example, the preset time interval... Let's take t+3t as the slope. At this point, the slope of the curve should reach m. In engineering, m is often set to a value slightly lower than the maximum torque response slope of the motor to ensure feasibility.

[0052] In step S323, the controller acquires the initial calibration values ​​of parameters t and a. The initial calibration value of t can be estimated based on the time window of gear grinding, for example, 50ms. The initial calibration value of a can be set based on the calibration value of a for a reference vehicle model. The reference vehicle model is a similar model to the vehicle whose n value is currently being calculated, i.e., a similar drive type, similar motor reducer, and similar rear axle structure. It can be understood that the calibration value of a for the reference vehicle model can be understood as historical data.

[0053] In step S324, the controller uses the motor's torque response slope as the target slope m, and substitutes the initial calibration values ​​of a and t into Formula 1 to form a nonlinear equation about n. Since n appears simultaneously in both the coefficient and exponent positions, it cannot be solved analytically. Therefore, the controller uses numerical methods, such as Newton's iteration method, to calculate an approximate value for n. For example, taking t = 0.05s, t0 = 0.2s, m = 600 N·m / s, and a = 2000, the equation becomes... Through iterative solution, n≈2.8 is obtained, and the controller is rounded to 3.

[0054] In this embodiment, by obtaining the motor's torque response slope, a target slope to be reached at a preset time after the zero-crossing time node in the target displacement power function curve is defined. Initial calibration values ​​for t and a are obtained, and the motor's torque response slope, as the target slope and initial calibration value, are substituted into the nonlinear equation to solve for n, establishing a quantitative mapping relationship between the motor's response capability and the power function exponent. This scheme ensures that the value of n is no longer an empirical tuning result, but rather an analytical calculation based on the motor's physical characteristics, significantly improving the adaptability of the control strategy to different motor platforms and greatly reducing the number of calibration tests.

[0055] In one embodiment of this application, S300 further includes, namely, adjusting the delay response time parameter, the rate of change parameter, and the trend of change parameter to make a preset curve segment of the displacement power function curve approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve, further comprising: S330, after t and n are determined, a is adjusted through a finite number of iterative experiments, and noise, vibration and acoustic roughness data corresponding to each experiment are collected. The value of a that minimizes the noise, vibration and acoustic roughness data is selected as the calibration value of a.

[0056] Specifically, in step S330, after parameters t and n have been determined according to the aforementioned method, the controller further refines the calibration of parameter a. The calibration process employs a finite number of iterative experiments: on a real vehicle or test bench, a is set to a series of different values, for example, between 0.5 times the estimated a value and 2 times the estimated a value, with 5 to 10 a values ​​taken at equal intervals. For each a value, a torque zero-crossing operation is performed, and NVH data is collected using NVH sensors mounted on the vehicle body. The NVH data comprehensively reflects the impact level during gear meshing and the driver's comfort. The controller establishes a correspondence between the collected NVH data and the a value, selecting the a value that minimizes the NVH data as the final calibration value. For example, if the experiment shows that the comprehensive NVH index is 0.3 when a=1500, 0.25 when a=2000, and 0.4 when a=2500, then a=2000 is selected as the calibration value for a.

[0057] In this embodiment, after t and n are determined, 'a' is adjusted through a finite number of iterative experiments, and noise, vibration, and acoustic roughness data corresponding to each experiment are collected. The 'a' value that minimizes this data is selected as the calibration value, thus achieving objective optimization of the rate of change parameter. This scheme transforms subjective driving experience into quantifiable noise and vibration indicators, making the parameter calibration results repeatable and verifiable, while ensuring the optimality of the final control strategy in terms of comfort.

[0058] In one embodiment of this application, S300 further includes, by adjusting the delay response time parameter, the rate of change parameter, and the trend of change parameter to make a preset curve segment of the displacement power function curve approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve, the method further includes: S340, after t and n are determined, a is adjusted through a finite number of iterative experiments, and motor speed fluctuation data corresponding to each experiment is collected. The value of a that minimizes the fluctuation peak of the motor speed fluctuation data is selected as the calibration value of a.

[0059] Specifically, in step S340, for vehicle platforms without NVH sensors, an alternative calibration method for parameter a is provided.

[0060] In this embodiment, after t and n are determined, the controller sets 'a' to a series of different values. For each 'a' value, a torque zero-crossing operation is performed, and the motor speed change curve over time is collected in real time by a speed sensor. When gear grinding occurs, a momentary spike fluctuation will appear on the speed curve. The controller extracts the maximum peak value of the speed fluctuation in each test and selects the 'a' value that minimizes the amplitude of the spike as the calibration value. For example, if the speed spike is 15 rpm when a=1800, 12 rpm when a=2200, and 20 rpm when a=2600, then a=2200 is selected. This method does not require additional sensors and can complete the optimization using only the existing speed signal from the motor controller.

[0061] like Figure 7 As shown, Figure 7 The vertical axis of the blue line represents the vehicle's requested torque, measured in N·m (Newton-meter). The vehicle's requested torque can be understood as how much torque the vehicle wants when the accelerator pedal is pressed. Figure 7 The vertical axis of the green line represents the motor response torque, measured in N·m (Newton-meter). Motor response torque can be understood as the actual torque value output by the motor controller (MCU, i.e., motor control unit) after receiving a "request torque" command, taking into account the motor's current actual capabilities (such as speed, temperature, available current, and bus voltage). Figure 7 The vertical axis, represented by the red line, shows the motor speed in rpm (revolutions per minute). The horizontal axis represents time; 10:06:37.2000 means 10:06:37.2 seconds.

[0062] like Figure 7 As shown, we need to find the point where the motor speed first jumps after the motor response torque changes its polarity at zero crossing from the time-motor speed curve. We consider the tooth-scratching phenomenon to occur in the next unit of time corresponding to this point, that is, at the moment after the first jump. If it is convenient to calculate the fluctuation peak of the motor speed fluctuation data, it can be understood that the tooth-scratching phenomenon occurs infinitely close to the point where the first jump occurs.

[0063] Figure 7 It's the process of torque changing from negative to positive. After crossing zero, the motor's response torque gradually increases from 0 to a positive value. We only focus on the point where the torque first jumps after crossing zero. This point is... Figure 7 Point E in the middle.

[0064] Optionally, if the trend after the first jump point is flat, and there are no subsequent jump points within a preset time period (e.g., within 1 second) after the first jump point, the vertical coordinate value of the first jump point is taken from the time-motor speed curve as the speed fluctuation value.

[0065] Optionally, if at least one subsequent jump point occurs within a preset time period after the first jump point, then the number J of subsequent jump points occurring within the preset time period is obtained, and the speed fluctuation value is calculated using the following formula 2.

[0066] Formula 2. Wherein, This represents the speed fluctuation value. It is the absolute value of the motor speed at the jump point that has the largest absolute value among all jump points, including the point where the first jump occurs and all subsequent jump points. J represents the absolute value of the motor speed at the transition point with the smallest absolute motor speed among all transition points, including the point where the first transition occurs and all subsequent transition points. J is the number of subsequent transition points.

[0067] Formula 2 signifies that when multiple speed jumps occur, the fluctuations of all jump points should be considered and averaged to observe the motor speed fluctuations. Subsequent jump points are defined as those occurring within a preset time period after the first jump. This is because only when subsequent speed jumps occur consecutively within a short period after the first jump can subsequent jumps be considered to be caused by the first jump. In this case, Formula 2 is needed to calculate the average fluctuation, making the calculation more consistent with actual gear grinding situations.

[0068] For example Figure 7 Then the second jump point F appears. This is a jump point where the rotation speed changes to a decreasing trend. Point F can be understood as a result of point E, or as a pair of gears colliding and then bouncing apart. Figure 7 In this case, the speed fluctuation value = (absolute value of the ordinate of point E - absolute value of the ordinate of point F) divided by 2.

[0069] The speed fluctuation value refers to the peak value of the motor speed fluctuation data. The smaller the speed fluctuation value, the better the corresponding 'a' value, and the more likely it is to be set as the calibration value of 'a'.

[0070] In this embodiment, after t and n are determined, 'a' is adjusted through a finite number of iterative experiments, and motor speed fluctuation data corresponding to each experiment is collected. The value of 'a' that minimizes the fluctuation peak is selected as the calibration value, providing another objective approach to parameter optimization. The motor speed fluctuation peak directly reflects the instantaneous impact during gear meshing. Using this as an optimization target can more accurately suppress gear breakage and is not affected by the placement of external noise sensors, making it suitable for vehicle platforms without NVH sensors.

[0071] In one embodiment of this application, the complete process of the target output torque changing from a first polarity across zero to a second polarity includes a torque unloading process in which the target output torque changes from a positive value through zero to a negative value.

[0072] The complete process of the target output torque changing from the first polarity across zero to the second polarity also includes the torque loading process of the target output torque changing from a negative value through zero to a positive value.

[0073] Specifically, the torque zero-crossing process includes two physically opposite operating conditions. The first is the torque unloading process, where the target output torque gradually decreases from a positive value, passing zero and becoming negative. This condition corresponds to the driver releasing the accelerator pedal or pressing the brake pedal, switching the vehicle from a driving state to an energy recovery or braking state. During this process, the gears transition from driving-side tooth surface contact to braking-side tooth surface contact, and the tooth clearance needs to be eliminated again. The second is the torque loading process, where the target output torque gradually increases from a negative value, passing zero and becoming positive. This condition corresponds to the driver pressing the accelerator pedal to start or accelerate, switching the vehicle from a coasting or reversing state to a driving state. Although the two processes are opposite in direction, both can be smoothly controlled using the displacement power function approximation method described in this application. For example, the unloading process can be configured by setting the target displacement power function curve as... , where 'a' can be negative or processed using its absolute value. The loading process uses a positive value 'a'. The controller automatically selects the corresponding processing logic based on the direction of the currently received instruction.

[0074] In this embodiment, by defining the complete process of the target output torque changing from the first polarity across zero to the second polarity as including the torque unloading process from a positive value crossing zero to a negative value and the torque loading process from a negative value crossing zero to a positive value, it is clarified that this method is applicable to two typical operating conditions: releasing the accelerator or regenerating braking energy, and starting or accelerating. This scheme enables the vehicle to achieve a smooth zero-crossing transition in different driving scenarios, avoiding the jerking sensation caused by the change in torque direction, and improving driving comfort under all operating conditions.

[0075] In one embodiment of this application, the motor torque zero-crossing control method further includes: S010, Obtain the motor torque demand curve. The horizontal axis of the motor torque demand curve represents time, and the vertical axis represents the required motor torque.

[0076] The custom curve also includes at least one other curve segment besides the smooth target torque curve segment, and the shape of the at least one other curve segment is the same as the curve segment in the corresponding time interval of the motor torque demand curve.

[0077] Specifically, S010 can be executed before S100.

[0078] In step S010, the controller receives the motor torque demand curve from the vehicle controller, i.e. Figure 6The L1 curve in the diagram represents time on the horizontal axis and the required torque of the motor on the vertical axis. This curve reflects the change in the target torque required by the driver's intention or the vehicle's energy management strategy over time. It is usually represented as a broken line that changes approximately as a straight line over time, or simply a straight line. For example, under rapid acceleration, the required torque may jump from 0 to 200 N·m in 0.1 seconds, forming a steep upward straight line.

[0079] In this embodiment, the custom curve not only includes the smooth target torque curve segment (the gradually changing segment in the zero-crossing interval), but also includes at least one other curve segment, such as... Figure 4 As shown, the smooth target torque curve segment is Figure 4 In the AB segment, the portion consisting of all points with an x-coordinate less than A is the other curve segment. Similarly, the portion consisting of all points with an x-coordinate greater than B is also an other curve segment. These other curve segments are located outside the time interval of the smooth target torque curve segment, i.e., before and after the zero-crossing process. The shape of these other curve segments is set to be exactly the same as the shape of the motor torque demand curve obtained in step S500 within the corresponding time interval. That is, only within the narrow interval of torque zero crossing does the custom curve adopt a smooth power function shape. In regions far from zero, the custom curve completely follows the original motor demand torque to ensure rapid response. For example, if the original demand torque drops linearly from +100Nm to -50Nm at t=0.5 seconds, and the zero-crossing interval is set to 0.5 seconds to 0.55 seconds, then before 0.5 seconds and after 0.55 seconds, the custom curve coincides with the original demand curve. Only between 0.5 seconds and 0.55 seconds is the custom curve replaced by a smooth power function transition segment.

[0080] Optionally, retrieving the target displacement power function curve in S400 and outputting the real-time torque value based on the target displacement power function curve means outputting the real-time torque value according to the target displacement power function curve during the stage of changing from the first polarity crossing zero to the second polarity, that is, during the zero-crossing interval... Figure 6 L3 in the code outputs real-time torque. During periods when the torque polarity does not need to cross zero, the real-time torque value is output based on the motor torque demand curve. That is, when actually outputting the torque value, i.e., when executing S400, only the CD segment is limited to... This formula outputs the real-time torque value. That is, segment CD outputs the real-time torque value according to L3, while other segments are consistent with L1, that is, other segments output the real-time torque value according to L1. This formula only applies to the intervals that cross zero.

[0081] In this embodiment, by obtaining the motor torque demand curve and defining that the custom curve further includes at least one other curve segment in addition to the smooth target torque curve segment, and the at least one other curve segment has the same shape as the curve segment in the corresponding time interval of the motor torque demand curve, it is clarified that this method only performs smoothing processing in the zero-crossing interval, and directly follows the original demand torque in the large torque interval outside the zero-crossing interval. This solution avoids unnecessary response delays, ensures the fast response ability of the motor during large torque output, and achieves the balance between smoothness and responsiveness.

[0082] In one embodiment of the present application, n is greater than 1.

[0083] Specifically, the parameter n is the exponent in the displacement power function, and its numerical value determines the bending shape of the curve and the initial slope at the zero-crossing point. When n > 1, the first derivative (slope) of the function at x = t is zero. This means that when the torque starts to change from zero, the initial change rate is zero, and the torque increases very slowly from zero, which fully meets the physical requirement of avoiding impact at the initial stage of gear meshing. Taking n = 3 as an example, the slope is 0 at x = t, and the slope gradually increases as x increases, realizing the smooth feature of "continuously increasing from zero". If n = 1, the slope is a constant a, and there is a corner at the zero-crossing point, still with a slight impact. If 0 < n < 1, the slope approaches infinity at x = t, and the torque changes suddenly at the zero-crossing point, which instead exacerbates the tooth beating. Therefore, the present application defines n > 1, preferably an integer such as n = 2, 3, or 4. In practical applications, n = 3 (cubic parabola) is preferentially adopted due to its good symmetry and smoothness.

[0084] In this embodiment, by defining the exponent n to be greater than 1, it is ensured that the first derivative of the displacement power function at the zero point x = t is zero, so that the absolute value of the torque change rate at the starting moment of the target displacement power function curve passing through zero is zero. This technical feature is the mathematical prerequisite for realizing the smooth feature of continuously increasing from zero, excluding the cases where the initial slope is non-zero or infinite when n ≤ 1, and fundamentally ensuring the elimination effect of zero-crossing impact.

[0085] In one embodiment of the present application, the torque unloading process and the torque loading process respectively adopt the independently calibrated delay response time parameter, the change rate parameter, and the change trend parameter, and respectively generate a target displacement power function curve corresponding to the torque unloading process and a target displacement power function curve corresponding to the torque loading process.

[0086] Specifically, since the gear stress state, backlash direction, and driver's desired response speed differ under torque unloading (torque change from positive to negative) and torque loading (torque change from negative to positive) conditions, this application further optimizes the process by independently calibrating each process. Specifically, for the loading process, the controller calibrates a set of parameters (t_load, a_load, n_load) to generate the corresponding target displacement power function curve y_load. For the unloading process, the controller calibrates another set of parameters (t_unload, a_unload, n_unload) to generate the corresponding target displacement power function curve y_unload. The two sets of parameters can be completely different. For example, during loading, if the driver desires a responsive start, t_load can be set to a smaller value of 30ms, and a_load to a larger value to allow the torque to rise rapidly. During unloading, if the driver is more sensitive to jerking, t_unload can be set to a larger value of 60ms, and a_unload to a smaller value to allow the torque to decrease slowly. The controller automatically selects and calls the corresponding curve based on the type of instruction received. In the actual implementation, the controller stores two calibration tables, which correspond to the loading and unloading conditions respectively.

[0087] In this embodiment, by defining the torque unloading and torque loading processes separately using independently calibrated t, a, and n parameters and generating corresponding target displacement power function curves, the differences in gear force state, tooth backlash direction, and driver's desired response speed under loading and unloading conditions are fully considered. This scheme allows the unloading process to prioritize suppressing jerking while allowing for a slightly longer delay, while the loading process can balance smoothness and rapid response, achieving refined control under dual conditions and further improving driving comfort in all scenarios.

[0088] This application also provides a motor torque zero-crossing control system.

[0089] It should be noted that, for the sake of brevity, all hardware structures appearing in this application are labeled in the embodiments of the motor torque zero-crossing control system, and not in the embodiments of any of the aforementioned motor torque zero-crossing control methods.

[0090] like Figure 2 As shown, in one embodiment of this application, the motor torque zero-crossing control system includes a motor 10, a reduction gear set 20, a torque sensor 30, and a controller 40.

[0091] The reduction gear set 20 is mechanically connected to the output shaft 110 of the motor 10. The torque sensor 30 is used to detect the output torque or torque demand of the motor 10 in real time. There are one or more controllers 40. The controller 40 is communicatively connected to the torque sensor 30. The controller 40 is configured to execute the motor torque zero-crossing control method described in any of the foregoing embodiments.

[0092] When the controller 40 receives a control command that changes the required torque of the motor from the first polarity to the second polarity across zero, the controller retrieves the target displacement power function curve and outputs a real-time torque value according to the target displacement power function curve, so as to reduce the impact during the torque zero-crossing engagement of the reduction gear set 20.

[0093] Specifically, the motor torque zero-crossing control system consists of the following physical components: Motor 10, typically a permanent magnet synchronous motor or an AC asynchronous motor, serves as the driving power source.

[0094] The reduction gear set 20, whose input shaft is mechanically connected to the output shaft 110 of the motor 10 via a spline or coupling, is used to convert the high-speed, low-torque output of the motor 10 into a low-speed, high-torque output that is then transmitted to the wheels. The reduction gear set 20 contains multiple gear pairs with inherent backlash between them.

[0095] The torque sensor 30 can be installed between the output shaft 110 of the motor 10 and the input shaft of the reduction gear set 20, or integrated inside the motor 10, to collect in real time the actual torque value output by the motor 10 or the torque demand value received from the vehicle controller 40. The torque sensor 30 can be a strain gauge sensor or a magnetoelastic sensor.

[0096] One or more controllers 40, including a motor controller (MCU) and a vehicle controller (VCU). The controller 40 communicates with the torque sensor 30 via a CAN bus or a dedicated signal line. The controller 40 internally stores a computer program for a motor torque zero-crossing control method. Optionally, the motor torque feedback can also be directly read from the torque feedback value of the motor controller, without a specific torque sensor 30.

[0097] When the vehicle controller issues a command requiring torque to cross zero (e.g., from +50Nm to -30Nm), the motor controller immediately retrieves a pre-calibrated and stored target displacement power function curve (this curve has been approximated and optimized). The motor controller searches for the corresponding fitted output torque y on the curve based on the current time and sends this y value as a real-time torque command to the driver of motor 10, driving motor 10 to output the corresponding torque. Because the target displacement power function curve has an extremely low initial slope and a configurable delay time near the zero-crossing point, the gears engage smoothly during torque increase, significantly reducing shock and noise.

[0098] In this embodiment, a closed-loop control system comprising a motor 10, a reduction gear set 20, a torque sensor 30, and a controller 40 is constructed. The controller 40 is configured to execute the aforementioned motor torque zero-crossing control method, achieving complete hardware and software collaboration from custom curve generation and power function approximation to real-time torque output. This motor torque zero-crossing control system, without requiring additional mechanical structures or sensors, leverages the existing computing power of the controller 40 and significantly improves gear meshing impact through software-level algorithm upgrades. It possesses high engineering practicality and promotional value, ensuring stable and reliable reduction of zero-crossing gear breakage during actual vehicle operation.

[0099] The technical features of the above embodiments can be combined arbitrarily, and the execution order of the method steps is not restricted. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0100] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for controlling the zero-crossing torque of a motor, characterized in that, include: Generate a custom curve; the custom curve includes a smooth target torque curve segment, which corresponds in time to the complete process of the target output torque changing from the first polarity across zero to the second polarity; Construct a power function of displacement , where y is the fitted output torque, x is time, t is the delay response time parameter, a is the rate of change parameter, n is the trend of change parameter, and the curve corresponding to the displacement power function is defined as the displacement power function curve; By adjusting the delay response time parameter, the rate of change parameter, and the trend of change parameter, a preset curve segment of the displacement power function curve is made to approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve. When a control command is received that the motor's required torque changes from the first polarity across zero to the second polarity, the target displacement power function curve is retrieved, and the real-time torque value is output based on the target displacement power function curve.

2. The motor torque zero-crossing control method according to claim 1, characterized in that, The step of adjusting the delay response time parameter, the rate of change parameter, and the trend of change parameter to make a preset curve segment of the displacement power function curve approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve, includes: t is calibrated as the delay response time after the torque crosses zero, so that after waiting for time t after crossing zero, the real-time torque value is output according to the power function curve of the target displacement, and t is greater than 0; n is calculated based on the motor's torque response slope, and n is used to control the rate of change of the target output torque to be achieved after the preset curve segment ends.

3. The motor torque zero-crossing control method according to claim 2, characterized in that, The step of calculating n based on the motor's torque response slope, where n is used to control the rate of change of the target output torque to be achieved after the preset curve segment ends, includes: Obtain the torque response slope of the motor; The target slope is defined as the slope that is reached at a preset time after the zero-crossing point in the target displacement power function curve. Obtain the initial calibration values ​​of t and a; The target slope is the torque response slope of the motor, the initial calibration value of t is taken as the value of t, and the initial calibration value of a is taken as the value of a. The nonlinear equation is determined by solving the nonlinear equation, as shown in Formula 1. Official 1; Where n is the trend parameter, a is the rate of change parameter, m is the target slope, and t is the response time parameter. The preset time after the zero-crossing time node in the target displacement power function curve.

4. The motor torque zero-crossing control method according to claim 1, characterized in that, The step of adjusting the delay response time parameter, the rate of change parameter, and the trend of change parameter to make a preset curve segment of the displacement power function curve approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve, further includes: After t and n are determined, a is adjusted through a finite number of iterative experiments, and noise, vibration and acoustic roughness data corresponding to each experiment are collected. The value of a that minimizes the noise, vibration and acoustic roughness data is selected as the calibration value of a.

5. The motor torque zero-crossing control method according to claim 1, characterized in that, The step of adjusting the delay response time parameter, the rate of change parameter, and the trend of change parameter to make a preset curve segment of the displacement power function curve approximate the smooth target torque curve segment, thereby obtaining the target displacement power function curve, further includes: After t and n are determined, a is adjusted through a finite number of iterative experiments, and motor speed fluctuation data corresponding to each experiment is collected. The value of a that minimizes the peak of the motor speed fluctuation data is selected as the calibration value of a.

6. The motor torque zero-crossing control method according to claim 1, characterized in that, The complete process of the target output torque changing from the first polarity across zero to the second polarity includes the torque unloading process of the target output torque changing from a positive value through zero to a negative value; The complete process of the target output torque changing from the first polarity across zero to the second polarity also includes the torque loading process of the target output torque changing from a negative value through zero to a positive value.

7. The motor torque zero-crossing control method according to claim 1, characterized in that, Also includes: Obtain the motor torque demand curve; the horizontal axis of the motor torque demand curve is time, and the vertical axis of the motor torque demand curve is the motor torque demand. The custom curve also includes at least one other curve segment besides the smooth target torque curve segment, and the shape of the at least one other curve segment is the same as the curve segment in the corresponding time interval of the motor torque demand curve.

8. The motor torque zero-crossing control method according to claim 1, characterized in that, n is greater than 1.

9. The motor torque zero-crossing control method according to any one of claims 1 to 8, characterized in that, The torque unloading process and the torque loading process respectively use independently calibrated delay response time parameters, change rate parameters and change trend parameters, and respectively generate target displacement power function curves corresponding to the torque unloading process and the torque loading process.

10. A motor torque zero-crossing control system, characterized in that, include: Electric motor; A reduction gear set is mechanically connected to the output shaft of the motor; A torque sensor is used to detect the output torque or torque demand of the motor in real time. One or more controllers, the controllers being communicatively connected to the torque sensor and configured to perform the motor torque zero-crossing control method as described in any one of claims 1 to 9; When the controller receives a control command that changes the required torque of the motor from the first polarity to the second polarity across zero, it retrieves the target displacement power function curve and outputs a real-time torque value based on the target displacement power function curve to reduce impact during the torque zero-crossing engagement of the reduction gear set.

Images