Feedforward compensation method and system based on acceleration feedback

Abstract

The invention discloses a feedforward compensation method and system based on acceleration feedback, and the method comprises the steps: obtaining and analyzing a motion instruction of a controller, determining expected speed information and expected acceleration information of a controlled object, determining an expected feedforward resultant force based on the expected speed information and the expected acceleration information, and determining a feedforward control signal based on the expected feedforward resultant force; determining the acceleration feedback force of the controlled object based on the real-time acceleration information of the controlled object; determining an acceleration feedback compensation force based on the acceleration feedback force, and determining an acceleration feedback compensation signal of the controlled object based on the acceleration feedback compensation force; and synthesizing the feedforward control signal and the acceleration feedback compensation signal to obtain a motion control instruction of the controlled object, and controlling the controlled object to move according to the motion control instruction. According to the method, open-loop feedforward prediction and closed-loop acceleration observation feedback are deeply fused, the robustness of external interference and internal reference change is enhanced, mechanical resonance is effectively inhibited, and the trajectory tracking precision, the anti-interference capability and the motion stability are improved.

Description

Technical Field

[0001] This invention relates to the field of mechanical control technology, and in particular to a feedforward compensation method and system based on acceleration feedback. Background Technology

[0002] In the field of high-speed, high-precision motion control, traditional control strategies are facing increasingly severe performance bottlenecks. Classical feedback control, represented by PID control, is essentially error-driven lag regulation. When actuators drive CNC machine tools to perform complex contour machining, or industrial robots complete high-speed pick-and-place operations, inherent inertia, friction, and mechanical flexibility lead to dynamic tracking errors. The feedback controller can only respond after an error occurs. This "make a mistake first, then correct it" approach inevitably results in significant trajectory lag and contour distortion in high-speed, high-acceleration motion scenarios, limiting further improvements in equipment production efficiency and machining accuracy. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a feedforward compensation method and system based on acceleration feedback, comprising: The motion commands from the controller are acquired and parsed to determine the desired velocity and acceleration information of the controlled object. The desired feedforward resultant force is determined based on the desired velocity information and desired acceleration information, and the feedforward control signal of the controlled object is determined based on the desired feedforward resultant force. Determine the real-time acceleration information of the controlled object, and determine the acceleration feedback force of the controlled object based on the real-time acceleration information; The acceleration feedback compensation force is determined based on the acceleration feedback force, and the acceleration feedback compensation signal of the controlled object is determined based on the acceleration feedback compensation force. The feedforward control signal and the acceleration feedback compensation signal are combined to obtain the motion control command of the controlled object, and the controlled object is controlled to move according to the motion control command.

[0004] Furthermore, the step of acquiring the motion commands from the controller and parsing the motion commands to determine the desired velocity and desired acceleration information of the controlled object includes: The motion commands from the controller are acquired, and the type of the motion command is determined to determine the parsing method, so as to determine the desired velocity and desired acceleration information of the controlled object. If the motion command is a discrete position command, the desired velocity and desired acceleration information are obtained through digital differential calculation. If the motion command is a continuous trajectory function, the desired velocity and desired acceleration information can be obtained by calculating the derivative of the function.

[0005] Furthermore, the step of determining the desired feedforward resultant force based on the desired velocity information and desired acceleration information, and determining the feedforward control signal of the controlled object based on the desired feedforward resultant force, includes: Determine the mass and viscous damping coefficient of the controlled object, and determine the desired velocity and desired acceleration of the controlled object from the desired velocity information and desired acceleration information; The desired feedforward resultant force of the controlled object is obtained by calculating based on the mass and viscous damping coefficient of the controlled object, as well as the desired velocity and desired acceleration. The desired feedforward resultant force is converted into a feedforward control signal for the force control mode, and this feedforward control signal is determined as the feedforward control signal for the controlled object.

[0006] Furthermore, the formula for calculating the expected feedforward resultant force of the controlled object is as follows: F = m*a + b*v Where F is the desired feedforward net force of the controlled object, m is the mass of the controlled object, a is the desired acceleration of the controlled object, b is the viscous damping coefficient of the controlled object, and v is the desired velocity of the controlled object.

[0007] Furthermore, determining the real-time acceleration information of the controlled object and determining the acceleration feedback force of the controlled object based on the real-time acceleration information includes: The real-time acceleration information of the controlled object is determined, and the real-time acceleration information is preprocessed, including bias removal, filtering and phase compensation, to obtain the real-time acceleration. The mass of the controlled object is determined, and based on Newton's second law, the real-time acceleration is multiplied by the mass of the controlled object to calculate the acceleration feedback force.

[0008] Furthermore, the step of determining the acceleration feedback compensation force based on the acceleration feedback force, and determining the acceleration feedback compensation signal of the controlled object based on the acceleration feedback compensation force, includes: The force opposite to the direction of the acceleration feedback force is determined as the acceleration feedback compensation force. The acceleration feedback compensation force is converted into the acceleration feedback compensation signal of the force control mode, and this acceleration feedback compensation signal is determined as the acceleration feedback compensation signal of the controlled object.

[0009] Furthermore, the step of synthesizing the feedforward control signal and the acceleration feedback compensation signal to obtain the motion control command for the controlled object, and controlling the controlled object to move according to the motion control command, includes: The motion control command of the controlled object is obtained by synthesizing the feedforward control signal and the acceleration feedback compensation signal. The motion control command of the controlled object is sent to the controller, and the controller controls the controlled object to move according to the motion control command.

[0010] Furthermore, the calculation formula for the motion control command of the controlled object is as follows: T = Tc + Tf + Ta Where T is the motion control command of the controlled object, Tc is the preset feedback correction force, Tf is the feedforward control signal, and Ta is the acceleration feedback compensation signal.

[0011] The present invention also provides a feedforward compensation system based on acceleration feedback, comprising: The acquisition module is used to acquire motion commands from the controller, parse the motion commands, and determine the desired velocity and desired acceleration information of the controlled object. The first determining module is used to determine the desired feedforward resultant force based on the desired velocity information and desired acceleration information, and to determine the feedforward control signal of the controlled object based on the desired feedforward resultant force. The second determining module is used to determine the real-time acceleration information of the controlled object and determine the acceleration feedback force of the controlled object based on the real-time acceleration information; The third determining module is used to determine the acceleration feedback compensation force based on the acceleration feedback force, and to determine the acceleration feedback compensation signal of the controlled object based on the acceleration feedback compensation force. The synthesis module is used to synthesize the feedforward control signal and the acceleration feedback compensation signal to obtain the motion control command of the controlled object, and to control the controlled object to move according to the motion control command.

[0012] Furthermore, the real-time acceleration information of the controlled object comes from the output of the acceleration sensor.

[0013] Compared with the prior art, the feedforward compensation method and system based on acceleration feedback of this invention have the following advantages: This invention introduces acceleration feedback, measures the actual acceleration in real time and converts it into a feedback force, and then generates a compensating force in the opposite direction. This compensating force can instantly cancel out the undesired acceleration components caused by model mismatch, sudden load changes, external disturbances, etc., which is equivalent to adding a considerable electronic damping to the system. This not only greatly enhances the robustness to external disturbances and changes in internal parameters, but also effectively suppresses mechanical structure resonance. This invention intelligently synthesizes a predictive feedforward control signal with an adaptive acceleration feedback compensation signal to form an optimal motion control command that combines speed and stability. This enables the controlled object to achieve extremely high trajectory tracking accuracy, strong anti-interference ability, and excellent motion stability during high-speed, high-acceleration motion. Attached Figure Description

[0014] Figure 1This is a schematic diagram of the flow structure of the feedforward compensation method based on acceleration feedback in an embodiment of the present invention; Figure 2 This is a schematic diagram of the composition of the feedforward compensation system based on acceleration feedback in an embodiment of the present invention. Detailed Implementation

[0015] The specific embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the invention.

[0016] like Figure 1 As shown in the embodiments of this application, a feedforward compensation method based on acceleration feedback is provided, including: S100: acquiring motion commands from the controller, parsing the motion commands, and determining the desired velocity information and desired acceleration information of the controlled object; S200: determining the desired feedforward resultant force based on the desired velocity information and desired acceleration information, and determining the feedforward control signal of the controlled object based on the desired feedforward resultant force; S300: determining the real-time acceleration information of the controlled object, and determining the acceleration feedback force of the controlled object based on the real-time acceleration information; S400: determining the acceleration feedback compensation force based on the acceleration feedback force, and determining the acceleration feedback compensation signal of the controlled object based on the acceleration feedback compensation force; S500: synthesizing the feedforward control signal and the acceleration feedback compensation signal to obtain the motion control command of the controlled object, and controlling the controlled object to move according to the motion control command.

[0017] Furthermore, this invention introduces acceleration feedback, which measures actual acceleration in real time and converts it into a feedback force, thereby generating a compensating force in the opposite direction. This compensating force can instantly cancel out the undesired acceleration components caused by model mismatch, sudden load changes, external disturbances, etc., which is equivalent to adding a considerable electronic damping to the system. This not only greatly enhances the robustness to external disturbances and changes in internal parameters, but also effectively suppresses mechanical structure resonance. This invention intelligently synthesizes the predictive feedforward control signal and the adaptive acceleration feedback compensation signal to form an optimal motion control command that combines speed and stability. This enables the controlled object to achieve extremely high trajectory tracking accuracy, strong anti-interference ability, and excellent motion stability during high-speed, high-acceleration motion.

[0018] In the embodiments of this application, a feedforward compensation method based on acceleration feedback is provided. The step of acquiring the motion command of the controller and parsing the motion command to determine the expected velocity information and expected acceleration information of the controlled object includes: acquiring the motion command of the controller and determining the parsing method based on the type of the motion command to determine the expected velocity information and expected acceleration information of the controlled object; if the type of the motion command is a discrete position command, the expected velocity information and expected acceleration information are obtained by digital differentiation calculation; if the type of the motion command is a continuous trajectory function, the expected velocity information and expected acceleration information are obtained by function differentiation calculation.

[0019] Specifically, the controller first receives motion commands from the higher-level planner and automatically determines their type. If the command is a discrete sequence of position commands, it is analyzed using digital differentiation, i.e., differential calculation is performed on the position commands at adjacent time points. If the command originates from a continuous trajectory function, it is analyzed using function differentiation, i.e., the position function is directly mathematically differentiated to obtain theoretically completely smooth and lag-free analytical expressions for the desired velocity and acceleration. This step ensures the quality of the state commands from the source. This technology lays a solid foundation for subsequent integrated feedforward and feedback control, thus significantly improving the overall trajectory tracking accuracy, dynamic response speed, and motion stability of the system. It is a core prerequisite for achieving high-speed and high-precision motion control.

[0020] In embodiments of this application, a feedforward compensation method based on acceleration feedback is provided. The method involves determining a desired feedforward resultant force based on desired velocity information and desired acceleration information, and determining a feedforward control signal for the controlled object based on the desired feedforward resultant force. The method includes: determining the mass and viscous damping coefficient of the controlled object, and determining the desired velocity and desired acceleration of the controlled object from the desired velocity information and desired acceleration information; calculating the desired feedforward resultant force of the controlled object based on the mass and viscous damping coefficient, as well as the desired velocity and desired acceleration; converting the desired feedforward resultant force into a feedforward control signal for a force control mode, and determining this feedforward control signal as the feedforward control signal for the controlled object.

[0021] Specifically, the desired velocity and desired acceleration are extracted from the parsed motion commands. The desired feedforward resultant force is calculated as the vector sum of inertial force and viscous damping force. This desired feedforward resultant force is directly mapped or converted into a feedforward control signal acceptable to the actuator in force control mode, thus completing the generation of the feedforward control signal. This step achieves predictive driving of the ideal dynamic response. Through feedforward calculation based on an accurate model, an ideal driving force that precisely overcomes the inherent inertia and damping of the system can be applied almost without delay as the motion commands change. This significantly eliminates the phase lag caused by waiting errors in traditional feedback control, fundamentally reducing dynamic tracking errors, especially in high-speed, high-acceleration motion segments.

[0022] In an embodiment of this application, a feedforward compensation method based on acceleration feedback is provided, wherein the formula for calculating the desired feedforward resultant force of the controlled object is: F = m*a + b*v Where F is the desired feedforward net force of the controlled object, m is the mass of the controlled object, a is the desired acceleration of the controlled object, b is the viscous damping coefficient of the controlled object, and v is the desired velocity of the controlled object.

[0023] In the embodiments of this application, a feedforward compensation method based on acceleration feedback is provided. The method for determining the real-time acceleration information of the controlled object and determining the acceleration feedback force of the controlled object based on the real-time acceleration information includes: determining the real-time acceleration information of the controlled object and preprocessing the real-time acceleration information, the preprocessing including bias removal, filtering and phase compensation to obtain the real-time acceleration; determining the mass of the controlled object and multiplying the real-time acceleration by the mass of the controlled object based on Newton's second law to obtain the acceleration feedback force.

[0024] Specifically, the real-time acceleration information of the controlled object is determined by using a high dynamic response accelerometer or by performing a second differentiation on the signal from a high-resolution position encoder. Since the original signal inevitably contains sensor bias, environmental vibration noise, and high-frequency interference introduced by the differentiation, it must undergo precise preprocessing. This includes bias removal to eliminate DC components and the influence of gravity, filtering to suppress high-frequency noise and resonant frequency interference, and phase compensation to correct signal lag and ensure timing accuracy, ultimately yielding an accurate and clean real-time acceleration signal. Combining this with the mass of the controlled object, and strictly adhering to Newton's second law, the processed real-time acceleration is multiplied by the mass to calculate the acceleration feedback force. This force, in its physical essence, represents the instantaneous net force required to generate the current actual acceleration. It comprehensively reflects all force effects acting on the system, including controller output, external disturbances, model errors, and unmodeled dynamics. This preprocessing step ensures the accuracy of the observation by providing a high-quality acceleration signal, while the calculation based on Newton's laws transforms the kinematic quantity of acceleration into the dynamic quantity of force, which directly corresponds to it, providing the controller with precise physical observations describing the actual force state.

[0025] In an embodiment of this application, a feedforward compensation method based on acceleration feedback is provided. The method for determining acceleration feedback compensation force based on acceleration feedback force and determining acceleration feedback compensation signal of the controlled object based on acceleration feedback compensation force includes: determining a force opposite to the direction of acceleration feedback force as acceleration feedback compensation force, converting acceleration feedback compensation force into acceleration feedback compensation signal of force control mode, and determining the acceleration feedback compensation signal as acceleration feedback compensation signal of the controlled object.

[0026] Specifically, the force opposite in direction to the calculated acceleration feedback force is determined as the acceleration feedback compensation force, forming a direct force negative feedback loop in the control principle: if the actual acceleration direction is the same as the desired direction (too large), a reverse compensation force is generated to suppress it; otherwise, a positive compensation force is generated to supplement it. This acceleration feedback compensation force is converted into an acceleration feedback compensation signal that matches the actuator force control mode through a linear transformation relationship, and is ultimately determined as the compensation command acting on the controlled object. This step, through the form of a force closed loop, directly unifies "observation" and "control" at the physical level, and can instantly counteract any undesired acceleration changes caused by model errors, load disturbances, or external shocks. It not only significantly improves the dynamic stiffness and anti-interference ability of the system, but also effectively suppresses mechanical resonance peaks and broadens the stability margin of the system. The compensation force works in conjunction with the feedforward force, allowing the feedforward controller to focus on achieving the ideal dynamic trajectory, while the acceleration feedback is specifically responsible for suppressing deviations. The two have clear division of labor and complement each other, jointly achieving a unity of high speed, high precision, and strong robustness.

[0027] In an embodiment of this application, a feedforward compensation method based on acceleration feedback is provided. The method involves synthesizing a feedforward control signal and an acceleration feedback compensation signal to obtain a motion control command for the controlled object, and controlling the controlled object to move according to the motion control command. The method includes: synthesizing and calculating the motion control command for the controlled object based on the feedforward control signal and the acceleration feedback compensation signal; sending the motion control command for the controlled object to a controller; and having the controller control the controlled object to move according to the motion control command.

[0028] Specifically, the predictive feedforward control signal and the adaptive acceleration feedback compensation signal are intelligently fused to generate the final motion control command for the controlled object. This command is sent to the underlying controller in real time, which drives the actuators according to the physical meaning of the command, thereby precisely controlling the controlled object to produce the desired actual motion. This step enables the actual motion trajectory of the controlled object to reproduce the desired complex dynamic command with high fidelity. It not only responds extremely quickly, following high-speed changing commands with almost no lag, but also has strong anti-interference and correction capabilities. When faced with load fluctuations, errors, or external disturbances, it can quickly maintain trajectory accuracy through real-time synthesized commands, and the smoothness and consistency of motion are greatly improved, effectively avoiding overshoot and oscillation.

[0029] In an embodiment of this application, a feedforward compensation method based on acceleration feedback is provided, wherein the calculation formula for the motion control command of the controlled object is: T = Tc + Tf + Ta Where T is the motion control command of the controlled object, Tc is the preset feedback correction force, Tf is the feedforward control signal, and Ta is the acceleration feedback compensation signal.

[0030] like Figure 2 As shown in the embodiments of this application, a feedforward compensation system based on acceleration feedback is provided, comprising: an acquisition module, configured to acquire motion commands from a controller, parse the motion commands, and determine the desired velocity information and desired acceleration information of the controlled object; a first determination module, configured to determine the desired feedforward resultant force based on the desired velocity information and desired acceleration information, and determine the feedforward control signal of the controlled object based on the desired feedforward resultant force; a second determination module, configured to determine the real-time acceleration information of the controlled object, and determine the acceleration feedback force of the controlled object based on the real-time acceleration information; a third determination module, configured to determine the acceleration feedback compensation force based on the acceleration feedback force, and determine the acceleration feedback compensation signal of the controlled object based on the acceleration feedback compensation force; and a synthesis module, configured to synthesize the feedforward control signal and the acceleration feedback compensation signal to obtain motion control commands for the controlled object, and control the controlled object to move according to the motion control commands.

[0031] In an embodiment of this application, a feedforward compensation system based on acceleration feedback is provided, comprising: real-time acceleration information of the controlled object originating from the output of an acceleration sensor.

[0032] In summary, this invention provides a feedforward compensation method and system based on acceleration feedback, comprising: acquiring and parsing the motion command of the controller, determining the desired velocity information and desired acceleration information of the controlled object, determining the desired feedforward resultant force based on the desired feedforward resultant force, and determining the feedforward control signal based on the desired feedforward resultant force; determining the acceleration feedback force of the controlled object based on the real-time acceleration information of the controlled object; determining the acceleration feedback compensation force based on the acceleration feedback force, and determining the acceleration feedback compensation signal of the controlled object based on the acceleration feedback force; synthesizing the feedforward control signal and the acceleration feedback compensation signal to obtain the motion control command of the controlled object, and controlling the controlled object to move according to the command. This invention deeply integrates open-loop feedforward prediction with closed-loop acceleration observation feedback, enhancing robustness to external disturbances and changes in internal parameters, effectively suppressing mechanical resonance, and improving trajectory tracking accuracy, anti-interference ability, and motion stability.

[0033] Finally, it should be noted that those skilled in the art can obviously make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims and their equivalents, this invention also intends to include these modifications and variations.

[0034] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention.

Claims

1. A feedforward compensation method based on acceleration feedback, characterized in that, include: The motion commands from the controller are acquired and parsed to determine the desired velocity and acceleration information of the controlled object. The desired feedforward resultant force is determined based on the desired velocity information and desired acceleration information, and the feedforward control signal of the controlled object is determined based on the desired feedforward resultant force. Determine the real-time acceleration information of the controlled object, and determine the acceleration feedback force of the controlled object based on the real-time acceleration information; The acceleration feedback compensation force is determined based on the acceleration feedback force, and the acceleration feedback compensation signal of the controlled object is determined based on the acceleration feedback compensation force. The feedforward control signal and the acceleration feedback compensation signal are combined to obtain the motion control command of the controlled object, and the controlled object is controlled to move according to the motion control command.

2. The feedforward compensation method based on acceleration feedback according to claim 1, characterized in that, The process of acquiring motion commands from the controller and parsing these commands to determine the desired velocity and desired acceleration information of the controlled object includes: The motion commands from the controller are acquired, and the type of the motion command is determined to determine the parsing method, so as to determine the desired velocity and desired acceleration information of the controlled object. If the motion command is a discrete position command, the desired velocity and desired acceleration information are obtained through digital differential calculation. If the motion command is a continuous trajectory function, the desired velocity and desired acceleration information can be obtained by calculating the derivative of the function.

3. The feedforward compensation method based on acceleration feedback according to claim 2, characterized in that, The process of determining the desired feedforward resultant force based on desired velocity information and desired acceleration information, and determining the feedforward control signal of the controlled object based on the desired feedforward resultant force, includes: Determine the mass and viscous damping coefficient of the controlled object, and determine the desired velocity and desired acceleration of the controlled object from the desired velocity information and desired acceleration information; The desired feedforward resultant force of the controlled object is obtained by calculating based on the mass and viscous damping coefficient of the controlled object, as well as the desired velocity and desired acceleration. The desired feedforward resultant force is converted into a feedforward control signal for the force control mode, and this feedforward control signal is determined as the feedforward control signal for the controlled object.

4. The feedforward compensation method based on acceleration feedback according to claim 3, characterized in that, The formula for calculating the expected feedforward net force of the controlled object is: F = m*a + b*v Where F is the desired feedforward net force of the controlled object, m is the mass of the controlled object, a is the desired acceleration of the controlled object, b is the viscous damping coefficient of the controlled object, and v is the desired velocity of the controlled object.

5. The feedforward compensation method based on acceleration feedback according to claim 3, characterized in that, The process of determining the real-time acceleration information of the controlled object and determining the acceleration feedback force of the controlled object based on the real-time acceleration information includes: The real-time acceleration information of the controlled object is determined, and the real-time acceleration information is preprocessed, including bias removal, filtering and phase compensation, to obtain the real-time acceleration. The mass of the controlled object is determined, and based on Newton's second law, the real-time acceleration is multiplied by the mass of the controlled object to calculate the acceleration feedback force.

6. The feedforward compensation method based on acceleration feedback according to claim 5, characterized in that, The process of determining the acceleration feedback compensation force based on the acceleration feedback force, and determining the acceleration feedback compensation signal of the controlled object based on the acceleration feedback compensation force, includes: The force opposite to the direction of the acceleration feedback force is determined as the acceleration feedback compensation force. The acceleration feedback compensation force is converted into the acceleration feedback compensation signal of the force control mode, and this acceleration feedback compensation signal is determined as the acceleration feedback compensation signal of the controlled object.

7. The feedforward compensation method based on acceleration feedback according to claim 6, characterized in that, The process of synthesizing the feedforward control signal and the acceleration feedback compensation signal to obtain the motion control command for the controlled object, and controlling the controlled object to move according to the motion control command, includes: The motion control command of the controlled object is obtained by synthesizing the feedforward control signal and the acceleration feedback compensation signal. The motion control command of the controlled object is sent to the controller, and the controller controls the controlled object to move according to the motion control command.

8. The feedforward compensation method based on acceleration feedback according to claim 7, characterized in that, The calculation formula for the motion control command of the controlled object is as follows: T = Tc + Tf + Ta Where T is the motion control command of the controlled object, Tc is the preset feedback correction force, Tf is the feedforward control signal, and Ta is the acceleration feedback compensation signal.

9. A feedforward compensation system based on acceleration feedback, characterized in that, include: The acquisition module is used to acquire motion commands from the controller, parse the motion commands, and determine the desired velocity and desired acceleration information of the controlled object. The first determining module is used to determine the desired feedforward resultant force based on the desired velocity information and desired acceleration information, and to determine the feedforward control signal of the controlled object based on the desired feedforward resultant force. The second determining module is used to determine the real-time acceleration information of the controlled object and determine the acceleration feedback force of the controlled object based on the real-time acceleration information; The third determining module is used to determine the acceleration feedback compensation force based on the acceleration feedback force, and to determine the acceleration feedback compensation signal of the controlled object based on the acceleration feedback compensation force. The synthesis module is used to synthesize the feedforward control signal and the acceleration feedback compensation signal to obtain the motion control command of the controlled object, and to control the controlled object to move according to the motion control command.

10. A feedforward compensation system based on acceleration feedback according to claim 9, characterized in that, The real-time acceleration information of the controlled object comes from the output of the acceleration sensor.

Images