A micro steering engine control method based on adaptive instruction filter

Abstract

The application discloses a kind of micro steering engine control methods based on adaptive instruction filter, the method first utilizes extension state observer to estimate load torque in real time;Second, the normalized load estimate is mapped to the time constant of instruction filter by nonlinear function, the filter bandwidth is adaptively reduced with load increasing;Finally, the error control is carried out to filtered instruction and feedback signal, and combined with load estimate value to carry out feedforward compensation, and drive signal is generated jointly.The application makes micro steering engine keep fast response when light load, effectively suppresses overshoot and oscillation when heavy load or load mutation, significantly improves the dynamic performance and robustness of system in whole working condition, and engineering implementation is simple, and compatibility is good.

Description

Technical Field

[0001] This invention belongs to the field of high-precision servo control technology, specifically a micro servo motor control method based on an adaptive command filter. Background Technology

[0002] Miniature servo motors, as the core actuators of precision servo control systems, are widely used in high-performance motion control scenarios such as robot joint driving, aerospace model control, and intelligent equipment positioning. Their control performance directly affects the system's positioning accuracy, response speed, and operational stability.

[0003] Currently, micro servos generally employ a closed-loop position control architecture. This architecture generates an error signal by comparing the desired position command with the actual position feedback signal, and then uses linear controllers such as PID and PI to adjust this error to achieve position tracking. This classic control method is simple in structure, easy to implement, and exhibits good control performance in situations with constant load and gradual command changes.

[0004] However, in practical engineering applications, especially in situations with frequent load changes and significant command jumps, traditional control methods have revealed the following key technical problems:

[0005] 1. Control saturation and overshoot caused by large initial errors. When the initial error of the system is large, the output of the traditional PID controller will change abruptly, which can easily lead to saturation of the drive circuit. Once it enters the saturation region, the control system is actually in an open-loop state and cannot respond to error changes in a timely manner, thus causing significant overshoot and oscillation, which seriously affects the transient performance and stability of the system.

[0006] 2. Fixed-bandwidth command filters lack adaptability. To smooth command jumps and suppress high-frequency interference, first-order inertial filters (low-pass circuits) are often introduced into the command channel. However, the bandwidth (or time constant) of such filters is usually fixed and cannot be dynamically adjusted according to the actual load. Under light load or stable conditions, an excessively wide bandwidth may lead to slow response; while under heavy load or rapid maneuvering conditions, an excessively narrow bandwidth will limit the system response speed, causing tracking lag. Therefore, fixed-bandwidth filters are difficult to optimize comprehensively between speed, accuracy, and overshoot.

[0007] 3. Load variations have a significant impact on system performance. In actual operation, miniature servos often face changes in load inertia and external disturbances, resulting in significant alterations to their dynamic characteristics. Traditional fixed-parameter controllers and filters cannot adapt to these changes, often exhibiting significant under- or over-adjustment during sudden load changes, and even triggering continuous oscillations, thus reducing the system's robustness and control quality under different operating conditions.

[0008] In summary, existing micro servo control methods suffer from problems such as control saturation, difficulty in balancing response speed and overshoot suppression, and poor adaptability when dealing with load changes and sudden command changes. Therefore, there is an urgent need for an adaptive control method that can dynamically adjust control parameters, especially the command filter bandwidth, according to the actual load conditions, in order to achieve high-precision, fast-response, and low-overshoot control of micro servos across the entire operating range. Summary of the Invention

[0009] To address the shortcomings of existing micro servo control methods, namely the difficulty of fixed-parameter controllers and command filters adapting to load changes, leading to problems such as control saturation and overshoot under large initial errors, and the inability to simultaneously achieve both response speed and overshoot suppression, this invention provides a micro servo control method based on an adaptive command filter. The core objective of this method is to fundamentally resolve the aforementioned contradictions by sensing load changes in real time and dynamically adjusting the smoothness of the command, thereby achieving high-precision, fast-response, and low-overshoot control of the micro servo across the entire operating range.

[0010] The technical solution of this invention is as follows:

[0011] A micro servo motor control method based on an adaptive command filter includes the following steps:

[0012] Step 1: Acquire the current operating status of the micro servo motor in real time, and based on the mathematical model of the motor, use an extended state observer to estimate the load torque acting on the shaft end of the micro servo motor in real time. ;

[0013] Step 2: Establish an adaptive mapping relationship between the load and the bandwidth of the command filter in the micro servo control system, and calculate the time constant of the command filter in real time. ;

[0014] Step 3: Calculate the time constant obtained in step 2 in real time. Command filters used in micro servo control systems process the raw command signals. Smoothing is performed to obtain the filtered command signal. The filtered command signal The tracking error is compared with the actual feedback signal from the micro servo motor, and the resulting error is used by the controller to generate the first control signal; simultaneously, the estimated load torque obtained in step 1 is used... Feedforward compensation is performed to generate a second control signal; the first control signal and the second control signal are superimposed to drive the micro servo motor together.

[0015] In a further preferred embodiment, the mathematical model of the electric servo motor in step 1 is as follows:

[0016]

[0017] in, For servo angular displacement, For the servo motor angular velocity, For armature current, The torque constant is This refers to the moment of inertia of the servo motor shaft and the load, calculated on the same side. The viscous damping coefficient is... For load torque, The resistance of the armature winding of the servo motor. The inductance of the armature winding of the servo motor. The back electromotive force coefficient, This is the terminal voltage applied to the armature.

[0018] In a further preferred embodiment, in step 1, the extended state observer is:

[0019]

[0020] In the formula, and They are respectively and The estimated value, For the observer bandwidth, This represents the error between the ESO's observed and actual values ​​of the servo motor's angular velocity.

[0021] In a further preferred embodiment, step 2, the specific process of establishing the adaptive mapping relationship between the load and the command filter bandwidth in the micro servo control system is as follows:

[0022] The estimated load torque value obtained in step 1 Normalization is performed, and based on the normalized load estimate, the time constant of the command filter is calculated in real time using a preset nonlinear mapping function. The time constant Within the preset range [ τ min , τ max ] Internal changes.

[0023] In a further preferred embodiment, step 2 involves normalizing the estimated load torque value as follows:

[0024]

[0025] In the formula, The maximum operating load reference value is preset. This is the normalized load estimate.

[0026] In a further preferred embodiment, in step 2, the preset nonlinear mapping function is:

[0027]

[0028] in, A positive adjustable parameter used to control the steepness of the mapping curve; when hour, ;when hour, .

[0029] In a further preferred embodiment, in step 3, the command filter in the micro servo control system is a first-order inertial command filter:

[0030]

[0031] in This is either the original position command or the angle command. This is the filtered command signal. It is a time constant that changes in real time.

[0032] In a further preferred embodiment, in step 3, the controller is a PID controller or a state feedback controller.

[0033] A further preferred approach is that, in step 3, the load estimate is... Specifically, feedforward compensation involves adjusting the load estimate. Divide by torque constant The compensation signal is obtained and used as the second control signal. .

[0034] Beneficial effects

[0035] Compared with existing technologies, the micro servo control method based on adaptive command filters provided by this invention has the following significant advantages:

[0036] 1. This invention fundamentally solves the inherent contradiction that fixed-bandwidth filters cannot simultaneously achieve fast response and overshoot suppression by dynamically adjusting the command filter bandwidth in real time based on load estimation. The system responds quickly under light load and remains stable without overshoot under heavy load.

[0037] 2. The adaptive smoothing mechanism employed in this invention effectively avoids instantaneous saturation of the controller output caused by large step commands, suppressing the resulting oscillations at the source. Combined with load feedforward compensation, the system's robustness to external disturbances and load changes is significantly improved.

[0038] 3. The present invention is simple to implement and does not require changes to the basic framework of traditional servo control (PID + instruction filter). It only adds load observation based on ESO and a simple bandwidth scheduling algorithm, with low computational burden, easy to implement on existing microprocessor platforms, and high compatibility with mainstream servo firmware.

[0039] This invention achieves high-precision, fast-response, and low-overshoot control of micro servos across the entire operating range by driving adaptive command filter bandwidth scheduling through online load estimation and combining it with controller feedforward compensation. This significantly improves system stability and robustness and has broad engineering application prospects.

[0040] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0041] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0042] Figure 1 This is a model structure diagram of the adaptive filter, load estimation, and control law coupling of the present invention;

[0043] Figure 2 yes and A comparison chart of bandwidth variation curves;

[0044] Figure 3 It is the response curve of the servo motor from no load to load, without a command filter (load applied in 6s);

[0045] Figure 4 The servo motor switches from no-load to load, and the command filter has a fixed bandwidth. Response curve of =1 (load applied in 6s);

[0046] Figure 5 It is the response curve of the servo motor from no load to load, with the variable bandwidth of the command filter (load added in 6s). Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of the present invention clearer, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Those skilled in the art will understand that the following description is exemplary and not intended to limit the scope of protection of the present invention.

[0048] This invention proposes a micro servo control method based on an adaptive command filter. This adaptive filtering method dynamically adjusts the command filter bandwidth according to the actual load to balance speed, overshoot, and drive saturation limitations under different operating conditions. The method includes real-time dynamic scheduling of the command filter bandwidth through online load estimation, load-bandwidth mapping, and deep coupling of adaptive filtering with the control law. Specifically, it includes the following steps:

[0049] Step 1: Online load estimation:

[0050] This step utilizes the sensors built into the micro servo system, including position, speed, and current sensors, to acquire real-time system operating status information. Based on the mathematical model of the motor, an extended state observer (ESO) is used to estimate the load torque acting on the micro servo shaft in real time. .

[0051] Step 1.1: Establish the mathematical model of the electric servo motor:

[0052]

[0053] in, For servo angular displacement, For the servo motor angular velocity, For armature current, The torque constant is This refers to the moment of inertia of the servo motor shaft and the load, calculated on the same side. The viscous damping coefficient is... This is the load torque (with the direction of resistance to rotation being positive). The resistance of the armature winding of the servo motor. The inductance of the armature winding of the servo motor. The back electromotive force coefficient, This is the terminal voltage applied to the armature.

[0054] Step 1.2: Based on the mathematical model of the electric servo motor, design an extended state observer to estimate the load torque acting on the shaft end of the miniature servo motor in real time. .

[0055] According to the velocity loop motion equation in the state-space equation of the electric servo motor

[0056]

[0057] against Estimation is performed using an extended state observer (ESO):

[0058]

[0059] In the formula, and They are respectively and The estimated value, For the observer bandwidth, This represents the error between the ESO's observed and actual values ​​of the servo motor's angular velocity.

[0060] Step 2: Establish a mapping function between the load and the command filter bandwidth. This step maps continuous load estimates to continuous filter time constants.

[0061] For the traditional first-order inertial filter model:

[0062]

[0063] in This is either the original position command or the angle command. This is the filtered command signal. The time constant is the reciprocal of the bandwidth. A mapping function between the load and the command filter bandwidth is established through the following process, allowing the bandwidth to be adjusted in real time according to the load estimate. Under heavy load or rapid maneuvering conditions, the filter bandwidth is increased to improve response speed, while under light load or stable conditions, the bandwidth is decreased to suppress overshoot and chattering, achieving a balance between smooth command transition and high-frequency suppression.

[0064] Step 2.1: To prevent estimation noise interference and adapt to different load levels, the load estimate is normalized:

[0065]

[0066] In the formula, The maximum operating load reference value is preset. This is the normalized load estimate.

[0067] Step 2.2: Nonlinear mapping:

[0068] Let the filter time constant be in the interval [ τ min , τ max ] If the internal variation is considered, the mapping relationship between bandwidth (reciprocal of the time constant) and load estimate can be established as follows:

[0069]

[0070] when (Unloaded) ,have , The minimum time constant corresponds to the maximum filtering bandwidth (fastest response) under light load or no load conditions. This can be determined based on the maximum allowable initial step response overshoot of the system, thus reducing the smoothing effect of the command and ensuring the system's response speed while avoiding response lag caused by excessive smoothing. When fully loaded, there are , The maximum time constant corresponds to the minimum filtering bandwidth (strongest smoothing) under full load. It can be determined based on the system's stability time requirement for instruction step changes under full load, thereby enhancing the smoothing effect on instruction jumps and effectively suppressing overshoot and oscillation caused by increased inertia.

[0071] The shape adjustment parameter controls the overall "steepness" of the bandwidth variation curve, and . When the bandwidth is low, the transition is gradual, and the bandwidth increases evenly with the load; while when... When the load is large, the transition is steep, the bandwidth increases sharply under high load, and this means that once the load exceeds a certain threshold, the filter bandwidth will rapidly decrease to enhance smoothness. and For example, we will make a comparison, and the results are as follows: Figure 2 As shown, The value is usually selected in the range of [1,5] through simulation and experimentation to achieve a good trade-off between response speed and stability.

[0072] Step 3: Deeply couple the adaptive command filter, load estimation, and control law:

[0073] Step 3.1: Adaptive instruction filtering:

[0074] The calculation in step 2 will be performed in real time. Substitute into a first-order inertial filter

[0075]

[0076] In digital controllers (such as MCUs and DSPs), filtering is achieved using discretization methods such as backward differential.

[0077] Step 3.2: Feedback Control and Load Feedforward Compensation:

[0078] Filtered command The error is calculated by comparing the actual feedback signal with the input signal. The error is then input to the PID controller, which generates a command signal. Meanwhile, in order to proactively offset load disturbances, the load estimate is... After appropriate gain, such as This is converted into an equivalent instruction signal. Feedforward compensation is performed to further reduce steady-state error and overshoot; ultimately, the command signal is... and The superposition yields the final drive signal applied to the servo power amplifier.

[0079]

[0080] in This is the gain of the power amplifier. Additionally, the synthesized result needs to be adjusted. Amplitude limiting is applied to comply with hardware safety constraints.

[0081] Through the above method, a deeply coupled closed loop of load observation, bandwidth adaptive adjustment, command smoothing, feedback control, and load feedforward is achieved, enabling the control system to proactively adapt to load changes. The Simulink model of the controller structure after deep coupling of the adaptive filter, load estimation, and control law is shown below. Figure 1 As shown.

[0082] To evaluate the effectiveness of the micro servo control method based on the adaptive command filter designed in this invention, simulations were performed using MATLAB, with the following load conditions:

[0083] The motor is unloaded for 0-6 seconds, and a load is applied at the 6th second;

[0084] A motor model was built in MATLAB SIMULINK, and the simulation parameters were designed as follows:

[0085]

[0086]

[0087] PID controller parameters are set as follows: ;

[0088] Figure 2 For MATLAB simulation and The comparison chart of bandwidth variation curves shows that... The value of can control the steepness of the bandwidth variation curve. As the bandwidth increases, the curve becomes steeper, and the bandwidth surges dramatically.

[0089] Figure 3 , Figure 4 , Figure 5 The conditions are as follows: the motor is unloaded for the first 6 seconds, and then a load is applied in the 6th second; there is no command filter, and a fixed bandwidth command filter is applied. And the response curve with the addition of a variable bandwidth command filter. It can be seen that... Figure 3The system has a fast response time, but when the load is applied at the 6th second, it produces a significant overshoot (about 25%) and oscillations that last for several seconds. Figure 4 After adding a fixed-bandwidth command filter, the response is smooth and the overshoot is small under no-load conditions. However, when the load suddenly increases at the 6th second, an overshoot of about 15% still occurs, and the settling time is relatively long. Figure 5 A variable bandwidth command filter is added to the system. Under no-load conditions, due to load estimation... , The filter has a large bandwidth, and its response speed is close to that of the unfiltered case with no overshoot. At the moment of sudden load increase at the 6th second, the ESO quickly estimates the load increase. Rapidly increasing, leading to Adaptively increase to near within tens of milliseconds (like Figure 2 The filter bandwidth automatically decreases, resulting in a smoother response to command changes. The response curve shows a smooth load switching process, with overshoot essentially disappearing (<5%), and a faster response speed compared to a fixed filter.

[0090] This invention effectively resolves the contradiction between response speed and overshoot suppression in micro servos under varying load conditions by organically combining online load estimation, adaptive command filtering, and compensation, significantly improving the system's dynamic performance and robustness. The method of this invention only adds a load observer and adaptive filtering calculation module to the traditional PID control algorithm, resulting in low algorithm complexity, strong real-time performance, and ease of upgrade and implementation in existing servo control systems.

[0091] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention without departing from the principles and spirit of the present invention.

Claims

1. A micro servo motor control method based on an adaptive command filter, characterized in that: Includes the following steps: Step 1: Acquire the current operating status of the micro servo motor in real time, and based on the mathematical model of the motor, use an extended state observer to estimate the load torque acting on the shaft end of the micro servo motor in real time. ; Step 2: Establish an adaptive mapping relationship between the load and the bandwidth of the command filter in the micro servo control system, and calculate the time constant of the command filter in real time. The specific process for establishing the adaptive mapping relationship between the load and the command filter bandwidth in the micro servo control system is as follows: The estimated load torque value obtained in step 1 Normalization is performed, and based on the normalized load estimate, the time constant of the command filter is calculated in real time using a preset nonlinear mapping function. The time constant Within the preset range Internal changes; The preset nonlinear mapping function is: in, A positive adjustable parameter used to control the steepness of the mapping curve; when hour, ;when hour, ; Step 3: Calculate the time constant obtained in step 2 in real time. Command filters used in micro servo control systems process raw command signals. Smoothing is performed to obtain the filtered command signal. The filtered command signal The tracking error is compared with the actual feedback signal from the micro servo motor, and the resulting tracking error is used by the controller to generate the first control signal; simultaneously, the estimated load torque value obtained in step 1 is... Feedforward compensation is performed to generate a second control signal; the first control signal and the second control signal are superimposed to drive the micro servo motor together.

2. The micro servo control method based on an adaptive command filter according to claim 1, characterized in that: In step 1, the mathematical model of the electric servo motor is: in, For servo angular displacement, For the servo motor angular velocity, For armature current, The torque constant is This refers to the moment of inertia of the servo motor shaft and the load, calculated on the same side. The viscous damping coefficient is... For load torque, The resistance of the armature winding of the servo motor. The inductance of the armature winding of the servo motor. The back electromotive force coefficient, This is the terminal voltage applied to the armature.

3. The micro servo control method based on an adaptive command filter according to claim 2, characterized in that: In step 1, the extended state observer is: In the formula, and They are respectively and The estimated value, For the observer bandwidth, This represents the error between the ESO's observed and actual values ​​of the servo motor's angular velocity.

4. The micro servo control method based on an adaptive command filter according to claim 1, characterized in that: In step 2, the normalization process for the estimated load torque is specifically as follows: In the formula, The maximum operating load reference value is preset. This is the normalized load estimate.

5. The micro servo control method based on an adaptive command filter according to claim 1, characterized in that: In step 3, the command filter in the micro servo control system is a first-order inertial command filter: in This is either the original position command or the angle command. This is the filtered command signal. It is a time constant that changes in real time.

6. The micro servo control method based on an adaptive command filter according to claim 1, characterized in that: In step 3, the controller is a PID controller or a state feedback controller.

7. The micro servo control method based on an adaptive command filter according to claim 1, characterized in that: In step 3, the load estimate is... Specifically, feedforward compensation involves adjusting the load estimate. Divide by torque constant The compensation signal is obtained and used as the second control signal. .

Images