A control method for compensating for the nonlinear effect of the gear gap of an electric rudder

Abstract

The application discloses a control method for compensating for the nonlinear influence of the gear gap of an electric rudder. The method first performs offline identification of the gear gap parameters after power-on starting; then in the working state, the load side angle and the motor side angle are collected in real time, and the load side speed and acceleration are calculated by using a differential tracker to predict the tracking error of the next beat; then the load side given angle is input, and the motor side given angle is obtained according to the designed gear gap inverse model and the switching condition; finally, the appropriate given current is calculated through the motor position loop algorithm, so that the motor position follows the given motor position signal, and the tracking error of the electric rudder is reduced. The method provided by the application can effectively compensate for the steady-state error and tracking error caused by the gear gap, the algorithm is simple and easy to implement, and can also be applied to complex working conditions with acceleration and time-varying load.

Description

Technical Field

[0001] This invention relates to the field of electric servo motor servo control technology, and specifically to a control method for compensating for the nonlinear effects of gear backlash in electric servo motors. Background Technology

[0002] An electric servo motor is an integrated position servo drive widely used in military and civilian fields such as drones, missile control, robotics, and industrial control. Structurally, it typically consists of five parts: a motor, a gearbox, a control drive circuit, a sensor, and a housing. Traditional electric servos use a PWM demodulated voltage signal and a potentiometer voltage comparison to drive a DC motor for positioning. In recent years, permanent magnet synchronous motors (PMSMs) and their control technology have developed rapidly, naturally replacing DC motors as the power source for servos. Simultaneously, digital sensors and controllers have been introduced to replace analog servo systems. Compared to traditional servos, the new servos offer greater performance advantages and a wider range of applications, which also places higher demands on their control technology.

[0003] Electric servos, in addition to motor control, must also consider the nonlinear factors introduced by the gearbox, such as friction, dead zone, and backlash. These characteristics can lead to significant positioning and tracking errors, and in severe cases, even oscillation and loss of control. However, gears must have appropriate backlash during meshing to ensure a proper lubricating film between the tooth surfaces during operation and to prevent tooth jamming due to thermal expansion caused by temperature increases. Different types of reducers have different backlash sizes, and the treatment of backlash also differs. For example, for harmonic reducers, the backlash is negligible, but there is significant friction; while for reducers constructed with multi-stage gear pairs, backlash exists in each stage and is superimposed, therefore it cannot be ignored.

[0004] For backlash, a highly nonlinear characteristic, there are generally two approaches: mechanical backlash elimination and algorithmic compensation. Mechanical backlash elimination requires additional mechanical devices, resulting in high cost and system complexity; while algorithmic compensation for backlash is low-cost and widely applicable. Algorithmic compensation requires pre-modeling the backlash mechanism, and there are generally three modeling approaches. The first is the backlash model, which can intuitively reflect the backlash characteristics. Backlash compensation can be achieved by passing the input position signal through a dual inverse process. This inverse process is often called the backlash inverse model. However, the problem with this type of model is that it is based on a semi-closed-loop servo system, ignoring the load-side inertia and assuming that the load will instantly decelerate to 0 without motor torque, and the load can only move by relying on the motor. The second is the dead-zone model, which is based on a two-degree-of-freedom servo system model and can reflect the correspondence between torsional angle and torque, which is more consistent with the actual system situation. However, the model is heavily coupled, and the torque calculation requires high accuracy for system parameters such as stiffness, damping, and backlash width. Since these parameters are perturbed by conditions such as temperature, parameter identification algorithms are usually required, resulting in a relatively large computational load. The third type is the dead-zone model after smooth approximation. Although it can improve the torque impact caused by the dead-zone model, the position tracking performance will be significantly reduced. In addition, since smooth approximation requires calculations such as natural exponent and arctangent, the computational load is even greater, and the algorithm is more complex.

[0005] In summary, although current technologies can compensate for backlash to some extent under different conditions, they all have varying degrees of shortcomings, making it difficult to balance complexity and compensation effectiveness in engineering practice. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides a control method for compensating for the nonlinear effects of gear backlash in electric servo motors. This method is simple and easy to engineer, can effectively compensate for the influence of backlash characteristics, ensure the tracking performance and steady-state accuracy of the system, and is applicable to situations where acceleration and load vary over time.

[0007] This invention provides a control method for compensating for the nonlinear effects of gear backlash in an electric servo motor, comprising the following steps:

[0008] 1) The electric servo motor is modeled using a mathematical model of a two-degree-of-freedom servo system, and a switching function is introduced to construct the backlash inverse model of the electric servo motor gear.

[0009] 2) After powering on the electric servo motor, identify the backlash parameters of the servo gears, including the reciprocal of the gear ratio (m) and the backlash on the left side (c). l and the average width c on the right r Substitute the backlash parameters of the servo gear into the backlash inverse model;

[0010] 3) During operation, the load-side angle θ is collected in real time. l and motor side angle θ m The current load-side angle θ is tracked using a linear second-order differential tracker. l (k) obtains the current load-side velocity v les (k) and acceleration estimate a les (k); and predict the load-side angle θ of the next beat. les (k+1) and tracking error θ le (k+1);

[0011] 4) Input the given load position signal θ measured by the load-side position sensor into the backlash inverse model. ld (k), and simultaneously based on the current load-side angle θ l (k) Current motor side angle θ m (k) and tracking error θ le (k+1) Replace the conditions of the switching function and substitute the replaced switching function into the backlash inverse model to obtain the current given motor position signal θ. md (k);

[0012] 5) Calculate the given current i using the motor position loop. ref To make the motor side angle θ m Follow the given motor position signal θ md This reduces the tracking error of the electric servo motor.

[0013] As a preferred embodiment of the present invention, the switching function expression introduced in step 1) is:

[0014]

[0015] Where s is a logical expression;

[0016] The expression for the inverse model of the backlash of an electric servo motor gear is:

[0017] θ md (k)=θ ld (k) / m+χ rd (k)c r +χ ld (k)c l

[0018]

[0019] Where θ md (k) represents the given motor position signal, θ ld (k) represents the given load position signal; χ rd The switching function of the right branch in the backlash inverse model, χ ldLet be the switching function of the left branch in the backlash inverse model, and let χ be the value of . rd +χ ld =1;T gd This represents the desired reducer torque.

[0020] As a preferred embodiment of the present invention, the identification of tooth backlash parameters in step 2) specifically refers to:

[0021] The position signal of the triangular wave generated by the controller of the electric servo motor is tracked for n cycles, and the load-side angle θ is acquired and recorded in real time by the position sensor. l and motor side angle θ m , forming (θ m ,θ l A series of data points on the plane; then, linear regression is used to fit the rising and falling segments of each cycle, and the slopes of the rising and falling segments are denoted as m and m, respectively. ri ,m li (i = 1, 2, ..., n), its expression can be written as

[0022]

[0023] Where x is used to distinguish between the rising segment of r and the falling segment of l, i represents the period ordinal number, and k represents the sampling point ordinal number within each segment, with K points taken; the transmission ratio is averaged and rounded, and the expression is:

[0024]

[0025] Where [] denotes the rounding symbol. Then, the average right-side width c of the tooth gap is calculated from each ascending segment. ri The average width c on the left side of the tooth gap is calculated for the descending segment. li Its unified form can be written as:

[0026]

[0027] The final tooth backlash width is obtained by averaging the left and right widths calculated for each cycle:

[0028]

[0029] Among them, c l c is the average width on the left side of the tooth gap. r This represents the average width on the right side.

[0030] As a preferred embodiment of the present invention, in step 3), v les (k), a les (k), θ les (k+1), θ le The calculation method for (k+1) can be expressed as follows:

[0031] First, a second-order differential tracker is constructed, whose discrete system state equation is written as:

[0032]

[0033] The state variable x1(k) is used to track the current load-side angle θ. l (k), the state variable x2(k) is used to track the current load-side velocity v. l (k), the state variable x3(k) is used to track the current load-side acceleration a. l (k), where r is the linear filter factor, and T s Let v be the sampling period; let x2(k) be denoted as v les (k), x3(k) is denoted as a les (k), then the predicted value θ of the load-side angle in the next cycle. les The expression for (k+1) is:

[0034]

[0035] Tracking error θ le (k+1) takes into account the tracking delay of one clock cycle, and its expression is:

[0036] θ le (k+1)=θ ld (k)-θ les (k+1)

[0037] Where, θ ld (k) represents the given load position signal.

[0038] As a preferred embodiment of the present invention, the switching function expression in step (4) is replaced with:

[0039]

[0040] Where, θ l θ is the angle on the load side. m χ is the motor-side angle, ε is the tracking error threshold; rd χ is the switching function of the right branch in the backlash inverse model. ld Let be the switching function of the right branch in the backlash inverse model, where the switching rule ensures that when the given motor is on the left side of the backlash, χ rd =1, χ ld =0; when located to the right of the tooth gap, χ ld =1, χ rd =0; and χ always exists rd +χ ld =1.

[0041] As a preferred embodiment of the present invention, in step 5), the given current i is calculated through the motor position loop. ref The method involves designing using PID controllers, active disturbance rejection controllers, sliding mode controllers, fuzzy adaptive controllers, etc.

[0042] Compared with the prior art, the present invention brings the following beneficial gains:

[0043] (1) This invention proposes an inverse backlash model and its switching rules to compensate for backlash nonlinearity. The dead zone and near-dead zone models of backlash are highly coupled, making the control algorithm complex and parameter tuning difficult. The traditional backlash model is derived based on a semi-closed-loop electric servo system, ignoring the load-side inertia. Based on this ideal assumption, the model can only be applied under no-load, high-damping conditions. However, the backlash model corresponding to the inverse backlash model of this invention is derived based on the dead zone model and uses other feature information to replace the reducer torque calculation. This not only makes the algorithm simple and easy to implement, but also makes it applicable to complex working conditions where acceleration and load vary over time.

[0044] (2) This invention uses a second-order linear differential tracker to track the load angle, and then predicts the load angle and tracking error for the next cycle. This allows the effects of the current velocity and acceleration to be taken into account when calculating the tracking error, thus considering the delay error of one cycle. Applying this feature information makes the backlash model more sensitive in determining the compensation direction, thereby reducing the tracking error in situations such as commutation or sudden load changes. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of an electric servo motor.

[0046] Figure 2 This is a block diagram of the system control model.

[0047] Figure 3 (a) is a schematic diagram of the tooth gap model based on the gap model.

[0048] Figure 3 (b) is a schematic diagram of the inverse model of the tooth gap.

[0049] Figure 4 This is a block diagram of the system controller structure.

[0050] Figure 5 The following are the tracking waveforms of an electric servo motor using different backlash compensation strategies under load conditions, based on a sinusoidal given position signal; Figure 5 (a) under conditions without compensation; Figure 5 (b) Under the compensation strategy of the traditional backlash inverse model; Figure 5 (c) Under the backlash inverse model compensation strategy.

[0051] Figure 6The waveforms of an electric servo motor tracking a sinusoidal given position signal under no-load conditions, employing different backlash compensation strategies; among them Figure 6 (a) under conditions without compensation; Figure 6 (b) Under the compensation strategy of the traditional backlash inverse model; Figure 6 (c) Under the backlash inverse model compensation strategy. Detailed Implementation

[0052] To describe the present invention more specifically, the technology of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely for explaining the present invention and are not limited to the present invention.

[0053] Figure 1 This diagram illustrates the structure and connections of an electric servo motor, frequently used in miniaturized, low-speed, high-torque servo control applications. The structure integrates a motor, a cylindrical gear reducer, a motor-side position sensor, an output position sensor, and drive control circuitry. Both DC motors and PMSMs offer excellent control performance. When a surface-mount PMSM is used... d When vector control is applied with a voltage of 0, the q-axis voltage equation is the same as that of a DC motor. Here, we take a surface-mounted PMSM as an example for explanation.

[0054] like Figure 2 As shown, an electric servo motor can be modeled using a mathematical model of a two-degree-of-freedom servo system and decomposed into two parts: an electrical subsystem and a mechanical subsystem.

[0055] For electrical systems, the voltage equation and the mechanical equation can be written as follows:

[0056]

[0057]

[0058] Where, θ m For the mechanical angle of the motor; u q i q For the q-axis stator voltage and current; L, R, ψ f J m C m For the direct and quadrature axis inductance, stator resistance, and rotor flux linkage, the motor moment of inertia and viscous damping coefficient are given. p is the number of pole pairs of the motor. T e T g For electromagnetic torque, and for reducer torque, T. e Represented as

[0059]

[0060] The mechanical system is modeled using an inertial element with a dead zone. The dead zone is reflected in the calculation of the reducer torque, written as...

[0061]

[0062] Among them, K g Where m is the meshing stiffness, and c is the reciprocal of the transmission ratio. l c r c is the backlash parameter. r -c l Let Δθ be the backlash width. Δθ = mθ m -θ l θ l The load angle is given by equation (4). Equation (4) shows that the gear generates a torque due to torsional deformation, but does not transmit torque inside the tooth gap.

[0063] The load-side dynamic equation can be written as

[0064]

[0065] Among them, J l C l T represents the moment of inertia and viscous damping coefficient on the load side. l This represents the load torque.

[0066] If we assume the reducer is purely rigid, i.e., K g If the sum is +∞, then the modeling of the mechanical system can degenerate into a gap element, such as... Figure 3 As shown in (a). Written as

[0067]

[0068] like Figure 3 As shown in (b), to compensate for the adverse effects of the tooth gap, a direct approach is to construct an inverse model of the tooth gap. A switching function χ[s] is introduced to describe the inverse model of the tooth gap, which is a logical expression. When s is true, χ[s] = 1; otherwise, χ[s] = 0. The inverse model of the tooth gap can then be written as:

[0069] θ md (k)=θ ld (k) / m+χ rd (k)c r +χ ld (k)c l (7)

[0070]

[0071] Where, χ rd , χ ld Let be the switching function of the left and right branches in the backlash inverse model, and let χ be the value of . rd +χld =1. T gd This represents the desired reducer torque.

[0072] The gear backlash model and inverse backlash model were derived above based on a two-degree-of-freedom servo system model. However, two problems remain in application. First, the backlash width parameter is uncertain after each power-on, requiring online identification or offline identification after power-on. Second, during operation, T... gd It is difficult to determine directly, and other conditions need to be used to indirectly replace equation (8).

[0073] To address the aforementioned issues, the backlash parameters are first identified after power-on startup. The electric servo motor is then instructed to track a triangular wave position signal for n cycles, and the load-side angle θ is acquired and recorded in real-time using a position sensor. l and motor side angle θ m , forming (θ m ,θ l A series of data points on the plane. Then, linear regression is used to fit the rising and falling segments of each cycle, and the slopes of the rising and falling segments are denoted as m and m, respectively. ri m li (i = 1, 2, ..., n), its expression can be written as

[0074]

[0075] Where x distinguishes between the rising and falling segments, i represents the period ordinal number, and k represents the sampling point ordinal number within each segment, with K points selected. The transmission ratio is averaged and rounded, expressed as follows:

[0076]

[0077] Where [] denotes the rounding symbol. Then, the average right-side width c of the tooth gap is calculated from each ascending segment. ri The average width c on the left side of the tooth gap is calculated for the descending segment. li Its unified form can be written as

[0078]

[0079] The final tooth gap width is obtained by averaging the left and right widths of the tooth gap calculated in each cycle.

[0080]

[0081] Then, during operation, the load-side angle θ is collected in real time. l and motor side angle θ m Tracking θ using a linear second-order differential tracker l Obtain estimated values ​​of load-side velocity and acceleration v les ales Next, predict the load-side angle θ for the next cycle. les (k+1) and tracking error θ le (k+1).

[0082] The expression for the second-order differential tracker can be written as follows:

[0083]

[0084] Where x1 is used to track θ l x2 is used to track the load-side velocity v l x3 is used to track the load-side acceleration a l r is the velocity factor, which determines the filtering effect of the differential tracker. Assuming the acceleration remains constant within the sampling period of the position loop, the angle of the next cycle can be estimated using the formula for uniformly accelerated motion. Let x2 be denoted as v. les Let x3 be denoted as a les , then θ les The expression for (k+1) is

[0085]

[0086] Among them, t s The time interval for one cycle of the position loop. Position tracking error θ le (k+1) takes into account the tracking delay of one clock cycle, and its expression is:

[0087] θ le (k+1)=θ ld (k)-θ les (k+1) (15)

[0088] For θ le The calculation of (k+1) is for the design of switching conditions. Through θ le (k+1) and Δθ can reflect the directions of the given acceleration and load torque, respectively. le A large positive value (k+1) indicates that the given acceleration is positive, and the backlash compensation direction corresponds to the right; otherwise, it's on the left. When Δθ is on the right side of the backlash, it indicates that the load torque is to the left, and the backlash compensation direction corresponds to the right; otherwise, it's on the left. Based on the above analysis, the switching rule is redesigned as follows:

[0089]

[0090] Based on the backlash inverse model (7) and the switching condition (16), the given angle θ of the motor can be obtained. md Finally, the appropriate input current i is calculated using the motor position loop algorithm. qref Make the motor position θ m Follow θ mdThat's fine. For example... Figure 4 As shown, a method of nonlinear PD controller feedback and GPI state observer feedforward is provided, which can effectively realize motor position tracking.

[0091] First, a GPI observer is designed. During the motor's operation, it is subjected to frictional force that varies with speed and the torque of the reducer, which exhibits dead-zone characteristics. These are reflected in equation (2) as the damping term and T, respectively. g These forces exhibit strong nonlinearity and can be considered as lumped perturbations ξ(t), and equation (2) can be rewritten as follows:

[0092]

[0093] Wherein, the control input u = i q The disturbance term ξ(t) = -C m (dθ m / dt) / J m -T g Torque constant α = 1.5pψ f / J m ξ(t) can be estimated by designing a GPI observer. Assume ξ(t) (2) If ξ(t) is a bounded function in a very small neighborhood of 0, then the estimated value of ξ(t) is ξ. es (t) and its first derivative are used as state variables. The observer is designed in the form of...

[0094]

[0095] Where z1-z4 are state variables, and z1 and z2 are used to estimate θ. m and its first derivative, z3 is ξ es (t). e ξ =z1-θ m For θ m The estimation error is β1-β4, which is the observer gain. By properly configuring the observer gain, the GPI observer can be guaranteed to have good steady-state and dynamic performance.

[0096] The output of the disturbance observer is used as feedforward, and the nonlinear PD controller is used as feedback. The control law is designed as follows:

[0097]

[0098] Among them, K p K v The nonlinear feedback coefficient consists of a constant term and a nonlinear term related to the error, k p0 k p1 k p2 k v0 k v1k v2 Let be constant coefficients. The system error equation can be written as , or as .

[0099]

[0100] In the formula This represents the perturbation estimation error. In equation (20), although K... p K v For nonlinear coefficients, but K p The value of k p1 With k p1 +k p2 / k p0 Between, K v The value of k v1 With k v1 +k v2 / k v0 Between. Thus, through proper parameter tuning, the bounded parameter K... p K v Equation (20) can be made into a linearly stable perturbation system. At the same time, the design and parameter tuning of the GPI observer can ensure ξ e (t) is bounded, which gives the system bounded-input-bounded-output (BIBO) stability, and the error boundary is related to |ξ. e (t) is directly proportional to ξ. e When (t)| approaches 0, the tracking error will also approach 0.

[0101] The above implementation method was simulated in Simulink to obtain the results. Figure 5-6 .

[0102] Figure 5 The effect of an electric servo motor tracking a sine wave under different backlash compensation strategies under load conditions is shown. Figure 5 (a) represents the tracking effect under no compensation condition, where mc always exists. r Tracking error; Figure 5 (b) shows the tracking effect under the traditional backlash inverse model compensation. This model assumes that backlash is always compensated in the opposite direction to the motion. When the load is applied in the opposite direction to the motion, it can compensate for backlash. However, when the load is applied in the same direction, m(c) will appear instead. r -c l Tracking error; Figure 5 (c) shows the tracking effect under the backlash inverse model compensation, which can achieve fast and accurate tracking.

[0103] Figure 6 The effect of an electric servo motor tracking a sine wave under no-load conditions using different backlash compensation strategies is shown. Figure 6(a) Without compensation, an electric servo motor needs to pass through backlash before moving in either direction, thus creating mc in both directions. r and mc l Tracking error; Figure 6 (b) represents the compensation conditions for the traditional backlash inverse model. Figure 6 (c) represents the compensation conditions for the backlash inverse model; the overall tracking performance is good. (Explanation) Figure 6 (b) and Figure 6 (c) can be applied to no-load conditions.

[0104] The acceleration of the sine wave changes continuously at different stages. Simulation results under both loaded and unloaded conditions show that the proposed backlash compensation algorithm can quickly identify the backlash compensation direction and compensate accordingly, verifying the algorithm's effectiveness and applicability.

[0105] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention, but the present invention is not limited to the above embodiments. Any improvements and modifications made to the present invention by those skilled in the art based on the disclosure thereof should be within the scope of protection of the present invention.

Claims

1. A control method for compensating for the nonlinear effects of gear backlash in an electric servo motor, characterized in that, Includes the following steps: 1) A mathematical model of a two-degree-of-freedom servo system is used to model the electric servo motor, and a switching function is introduced to derive the backlash inverse model of the electric servo motor gears; the expression for the backlash inverse model of the electric servo motor gears is: in, Given a motor position signal, Given a load position signal; The switching function of the right branch in the backlash inverse model. Let be the switching function of the left branch in the backlash inverse model, and have ; Let m be the desired reducer torque, and m be the reciprocal of the transmission ratio. The average width on the left side of the tooth gap. The average width on the right side; 2) After powering on the electric servo motor, identify the backlash parameters of the servo gears, including the reciprocal of the gear ratio m and the average width of the left side of the backlash. and the average width on the right Substitute the backlash parameters of the servo gear into the backlash inverse model; 3) During operation, the load-side angle of the electric servo motor is collected in real time. and motor side angle The current load-side angle is tracked using a linear second-order differential tracker. Obtain the current load-side speed and acceleration estimates And predict the load-side angle of the next beat. and tracking error ; 4) Input the given load position signal measured by the load-side position sensor into the backlash inverse model. At the same time, based on the current load-side angle Current motor side angle and tracking error The conditions of the switching function are replaced, and the switched function with the replaced conditions is substituted into the backlash inverse model to obtain the current given motor position signal. ; The expression for the switching function after replacing the conditions is: in, For the load side angle, The angle is on the motor side. To track the error threshold, This is the switching function of the right branch in the backlash inverse model. Let be the switching function of the left branch in the backlash inverse model, and it always holds that... The switching rules ensure that when a given motor is on the left side of the backlash, , When located on the right side of the tooth gap, , ; 5) Calculate the given current using the motor position loop. To make the motor side angle Follow the given motor position signal This reduces the tracking error of the electric servo motor.

2. The control method for compensating for the nonlinear effects of gear backlash in an electric servo motor according to claim 1, characterized in that, In step 1) The introduced switching function expression is as follows Where s is a logical expression.

3. The control method for compensating for the nonlinear effects of gear backlash in an electric servo motor according to claim 1, characterized in that, The specific steps for identifying backlash parameters in step 2) are as follows: The position signal of the triangular wave generated by the controller of the electric servo motor is tracked for n cycles, and the load-side angle is acquired and recorded in real time by the position sensor. and motor side angle ,composition A series of data points on a plane; then, linear regression is used to fit the rising and falling segments of each cycle, and the slopes of the rising and falling segments are denoted as follows: Its expression can be written as Where x is used to distinguish between the rising segment of r and the falling segment of l, i represents the period ordinal number, and k represents the sampling point ordinal number within each segment, with K points taken; the transmission ratio is averaged and rounded, and the expression is: [] represents the rounding symbol; then the average width of the backlash on the right side is calculated from each ascending segment. The average width of the tooth gap on the left side is calculated for the descending segment. Its unified form can be written as: The average width of the tooth gap is obtained by averaging the widths of the left and right sides of the tooth gap calculated for each cycle. in, The average width on the left side of the tooth gap. This represents the average width on the right side.

4. The control method for compensating for the nonlinear effects of gear backlash in an electric servo motor according to claim 1, characterized in that, In step 3) , , , The calculation method can be expressed as: First, a second-order differential tracker is constructed, whose discrete system state equation is written as: Among them, state variables Used to track the current load-side angle State variables Used to track the current load-side speed State variables Used to track current load-side acceleration r is a linear filter factor. The sampling period; Recorded as , Recorded as Then the next load-side angle The expression for the predicted value is: Tracking error Considering the tracking delay of one frame, its expression is: in, Given the load position signal.

5. The control method for compensating for the nonlinear effects of gear backlash in an electric servo motor according to claim 1, characterized in that, In step 5), the given current is calculated using the motor position loop. The method involves designing a PID controller, an active disturbance rejection controller, a sliding mode controller, or a fuzzy adaptive controller.

Images