Multi-axis synchronous coordination control method based on virtual time-delay main axis

Abstract

The invention discloses a multi-axis synchronous coordination control method based on a virtual time-delay main axis, and the method comprises the following steps: S1, representing any axis in a multi-axis system with an x axis, building an observer, and obtaining the current position error Perr, observation speed Vx and observation acceleration Ax of the x axis; s2, the time delay time tdx corresponding to the x axis at the current moment is obtained through a formula; s3, for each axis in the multi-axis system, obtaining the time delay time corresponding to each axis at the current moment by adopting the method in the steps S1-S2; and S4, adopting a cross compensation control strategy, carrying out average calculation on the time delay time corresponding to each axis to obtain expected time delay time td, and superposing the expected time delay time td to a current loop regulator of the multi-axis system through a cross coupling regulator to realize cross compensation.

Description

Technical Field

[0001] The present invention relates to the technical field of synchronous coordinated control of multi-axis linkage, and in particular to a multi-axis synchronous coordinated control method based on a virtual time-delay main axis. Background Art

[0002] The key components of high-end equipment are often composed of complex curved surface parts, such as orthogonal variable ratio gears in various transmissions, large and complex molds, and turbine blades of aircraft engines. Their processing quality and effect are directly related to the performance and service life of high-end equipment. The production of complex curved surface parts depends on multi-axis linkage contour processing equipment, such as multi-axis precision CNC machine tools, industrial robotic arms, etc. As the core executive component of the multi-axis linkage control system, the performance of the servo system determines the processing quality and efficiency of the processing equipment. With the increase in the precision, complexity and demand of curved surface parts, processing equipment is also developing in the direction of high precision and high speed, and the impact of insufficient servo system performance has gradually become prominent. Taking advanced CNC machine tools as an example, the designed processing accuracy can reach IT8-IT5 or even higher, the fast tool feed speed can reach 120m / min, and the acceleration can reach more than 2g. However, the existing servo control performance is difficult to meet the contour processing accuracy, surface processing quality and processing efficiency at the same time. The insufficient dynamic response of the servo system and the lack of an ideal multi-axis coordinated control strategy make it impossible to guarantee the spatial contour tracking performance of the processing equipment, resulting in a decrease in processing accuracy as the processing speed increases, which ultimately affects the processing efficiency. In addition, due to the lack of adaptability of the servo system's software vibration suppression algorithm, it is unable to completely solve the problem of mechanical oscillation that is more likely to occur in high-response systems, which will also affect the processing quality of the equipment. Therefore, improving the control performance of multi-axis servo systems to meet the increasing demand for high-quality and efficient processing is a long-term task facing the modern manufacturing industry.

[0003] Complex surface processing equipment is essentially a multi-axis collaborative servo-driven mechanical system. The tracking characteristics and stability of the final spatial working curve of the multi-axis servo directly determine the processing effect of complex surface parts. Multi-axis coordination control strategies can be divided into two categories according to different control methods: one is the passive coordination control strategy of the servo system dominated by the host computer controller; the other is the active coordination control strategy dominated by the driver. In the passive coordination control strategy, each servo axis cannot obtain the operation information of other axes. The information of each axis is summarized to the host computer controller, and the host computer realizes synchronous control, such as multi-axis positioning coordination control. In this method, each axis follows the overall given signal separately, and there is no coupling between the axes. It is simple and intuitive, but the anti-disturbance performance is insufficient. In the active coordination control strategy, each axis uses high-speed communication networks such as industrial Ethernet to realize information transmission, obtain the operation data of other axes, and actively complete motion adjustment to achieve multi-axis synchronization.

[0004] Virtual spindle and cross-coupling are currently used multi-axis coordination control strategies. The virtual spindle synchronization control strategy follows the master-slave control structure, and the motion control board or host computer of the multi-axis system virtualizes a non-existent spindle. This method is simple to calculate and easy to implement, but it requires inherent synchronization between the axes to construct a virtual axis, which has limitations. The cross-coupling synchronization control strategy does not have the concept of a spindle, and each axis is in an equal position at the control level, which overcomes the limitations of the virtual axis solution. However, since the coupling relationship between a single axis and other axes needs to be considered, the complexity of the solution increases rapidly with the increase in the number of axes. Summary of the invention

[0005] In order to solve the problems raised in the above background technology, the technical solution adopted by the present invention is:

[0006] A multi-axis synchronous coordinated control method based on a virtual time-delay spindle comprises the following steps:

[0007] S1. For any axis in the multi-axis system, represented by the x-axis, an observer is established to obtain the current position error P of the x-axis. err , observation speed V x and the observed acceleration A x ;

[0008] S2, the time lag time t corresponding to the x-axis at the current moment is obtained by the following formula dx :

[0009]

[0010] S3, for each axis in the multi-axis system, the method described in steps S1-S2 is used to obtain the time lag time corresponding to each axis at the current moment;

[0011] S4, adopt the cross compensation control strategy, average the time delay time corresponding to each axis to obtain the desired time delay time t d , and is superimposed on the current loop regulator of the multi-axis system through the cross-coupling regulator to achieve cross compensation.

[0012] Furthermore, in step S1, when establishing the observer, the observer input includes the x-axis position given signal P xr The current signal C corresponding to the motor system feedback xf ;

[0013] The position given signal P xr Accept observation speed V x The integral of is used as negative feedback to form the position error P err ;

[0014] Position error P err After passing through the PID controller, the output acceleration signal a co, acceleration signal a co Considered to be caused by the resistance of the motor system;

[0015] Current signal C xf Multiply by the motor torque coefficient K t , the electromagnetic torque T can be obtained em , and then divided by the system inertia J, we can get the acceleration a generated by the electromagnetic torque e ;

[0016] Then the observed acceleration A x =a e +a co ;

[0017] Observed acceleration A x After integration, we get the observed velocity V x And participate in speed loop regulation.

[0018] Furthermore, after step S1 and before step S2, according to the obtained position error P err , observation speed V x and the observed acceleration A x To determine whether there is a situation where the time delay is unsolvable, if so, an error is reported to the system terminal and the process is terminated;

[0019] If there is no time delay without a solution, then determine the time delay t dx The positive and negative signs of the time lag time are as follows: a positive time lag time indicates that the current actual position lags behind the command position, and a negative time lag time indicates that the current actual position is ahead of the command position, and then the process goes to step S2.

[0020] Further, in step S2, specifically, the time lag time is solved by a root-finding formula of a quadratic equation;

[0021] For the two roots obtained, first determine whether the two roots meet the sign requirement based on the sign of the predetermined time delay;

[0022] If there is no root that meets the requirements, an error is reported to the system terminal and the process is terminated;

[0023] If there is a root that meets the requirements, it is used as the solved time delay;

[0024] If both roots meet the requirements, take the root with the smaller absolute value as the solved time lag.

[0025] Furthermore, in step S4, when cross compensation is performed, the control period of the cross compensation is made consistent with the calculation period of the current loop regulator, and data transparent transmission between the axes is achieved by means of high-speed industrial Ethernet.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] The multi-axis synchronous coordinated control method based on the virtual time-delay spindle provided by the present invention equates the response characteristics and operating status of each axis to the "time lag" time between the actual position and the command position, constructs a virtual time-delay spindle with the goal of consistent "time lag" of each axis, and then realizes the coupling compensation of each axis with the virtual time-delay axis, and finally coordinates the response performance of each axis to realize multi-axis synchronous control. Compared with the traditional solution, this solution has stronger universal applicability. It overcomes the limitation of the traditional virtual spindle solution that relies on inherent synchronization to build virtual axes; in actual application, each axis only needs to estimate the time lag time and coupling compensation value of the axis, and the complexity of the solution is independent of the number of system axes; the time lag time standardizes and normalizes the operating status of each axis, and the parameters of the compensation regulator do not need to be adjusted according to factors such as curve type and feed level. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1a-Figure 1c It is a graph of decomposed curves of each axis of the space trajectory curve C in a specific embodiment;

[0029] Figure 2 is a schematic diagram of an observer used in a specific embodiment;

[0030] Figure 3 It is a strategy block diagram of multi-axis synchronous coordinated control based on virtual time-delay main axis in a specific embodiment. DETAILED DESCRIPTION

[0031] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the following further describes how the present invention is implemented in conjunction with the accompanying drawings and specific implementation methods.

[0032] 1, the present invention provides a multi-axis synchronous coordinated control method based on a virtual time-delay spindle, comprising the following steps:

[0033] S1. For any axis in the multi-axis system, represented by the x-axis, an observer is established to obtain the current position error P of the x-axis. err , observation speed V x and the observed acceleration A x ;

[0034] S2, the time lag time t corresponding to the x-axis at the current moment is obtained by the following formula dx :

[0035]

[0036] S3, for each axis in the multi-axis system, the method described in steps S1-S2 is used to obtain the time lag time corresponding to each axis at the current moment;

[0037] S4, adopt the cross compensation control strategy, average the time delay time corresponding to each axis to obtain the desired time delay time t d , and is superimposed on the current loop regulator of the multi-axis system through the cross-coupling regulator to achieve cross compensation.

[0038] In a specific embodiment, it can be understood that in a multi-axis system, the motion control board or the host computer analyzes the position command of each axis according to the predetermined spatial trajectory in each control cycle, and the position command of each axis can be regarded as a function related to time, that is, the coordinates of each axis at any point on the spatial curve C can be represented by the time variable t, and all points of the spatial curve C can be obtained as t moves within a certain range. Taking a 5-axis system as an example, the trajectory functions of the x, y, z, a, and c axes can constitute the parametric equation of the spatial trajectory curve, as shown in the following formula:

[0039]

[0040] The decomposition curves of the space trajectory curve C(t) on each axis are C x , C y , C z , C a and C c , the curves correspond to the function x=f x (t), y = f y (t), z = f z (t), a=f a (t) and c = f c (t), refer to Figure 1a As shown in the figure, the curves of the z-axis and the a-axis are omitted. Assuming that the current time is t0, the spatial position of the composite of the axes, that is, the command position is C(t0) = {f x (t0),f y (t0),f z (t0),f a (t0),f c (t0)}. However, in practical applications, there must be a tracking error between the actual position of each axis at time t0 and the command position. Taking the x-axis as an example, if the x-axis motor moves according to the predetermined trajectory, it takes t d time to reach the command position at time t0, it can be equivalently considered that there is a t between the actual x-axis position and the command position. d The actual completion position is x = f x (t1), t1=t0-t d Therefore, if Figure 1a As shown in the figure, if the time lag between the actual position and the command position of each axis at time t0 is t d The spatial position of the actual arrival position of each axis is C(t1)={f x(t1),f y (t1),f z (t1),f a (t1),f c (t1)}, this point is still on the desired contour trajectory. If the lag time of each axis can be guaranteed to be consistent at any time, the trajectory tracking effect will be good.

[0041] However, if Figure 1b As shown in the figure, in actual applications, the response characteristics of each axis are not matched and the motion is not coordinated, resulting in different time delays. The actual time delay of each axis is t dx , t dy , t dz , t da and t dc , the spatial composite coordinates of the actual completed position at time t0 are {f x (t x1 ),f y (t y1 ),f z (t z1 ),f a (t a1 ),f c (t c1 )}, this point deviates from the space contour curve C(t).

[0042] Without considering the inter-axis coupling compensation, the time delay is related to the response characteristics and load state of the single axis: the higher the dynamic response of the single axis, the smaller the time delay, and vice versa; the time delay increases when the load suddenly increases, and vice versa. If the inter-axis coupling adjustment is considered, the system should control the axis with large time delay (slow response or sudden load change) to speed up the response and be in a catching-up state; while the axis with small time delay (fast response or sudden load change) should slow down the response and be in a waiting state, and finally ensure that the time delay of each axis is equal, so that the actual arrival position is on the spatial contour trajectory.

[0043] Reference Figure 1c As shown, if the desired time delay time t is obtained according to the method provided by the present invention d , and define the time axis corresponding to the expected time delay as the virtual time delay axis. The current time delay of the x-axis is t dx , the x-axis decreases the response and increases the x-axis lag along the red curve to t d ; The current time lag on the y-axis is t dy , then the y-axis increases the response, and the y-axis lag is reduced along the red curve to t d ; The final adjustment actual completion position is close to C(t1) = {f x (t1),f y (t1),f z (t1),f a (t1),fc (t1)}.

[0044] In order to obtain the desired delay time t d , the estimation of the single-axis time lag is crucial. Since the multi-axis coordinated control strategy further improves the multi-axis system contour tracking capability on the basis of single-axis position tracking, the single-axis position tracking error magnitude is within a controllable range when coordinated control is not added. In addition, the high-speed industrial bus increases the update rate of the position command, the system operation and adjustment cycle is very short (us level) and the mechanical inertia is relatively large. Therefore, the process of the single-axis system converging from the actual position to the command position in the current physical state can be approximately equivalent to an average acceleration process with a known initial velocity. Therefore, in step S1 of the present invention, for any x-axis, an observer is established to obtain the current position error P of the x-axis. err , observation speed V x and the observed acceleration A x Then, the time delay time t can be calculated by the formula in step S2. dx .

[0045] In this embodiment, in step S1, when establishing the observer, refer to Figure 2 As shown, the input of the observer includes the x-axis position given signal P xr The current signal C corresponding to the motor system feedback xf ; Wherein the position given signal P xr Accept observation speed V x The integral of is used as negative feedback to form the position error P err ; Position error P err After PID controller G co Output acceleration signal a co , acceleration signal a co Considered to be generated by the motor system resistance; current signal C xf Multiply by the motor torque coefficient K t , the electromagnetic torque T can be obtained em , and then divided by the system inertia J, we can get the acceleration a generated by the electromagnetic torque e ; Then the observed acceleration A x =a e +a co ; Observed acceleration A x After integration, we get the observed velocity V x And participate in speed loop regulation.

[0046] In addition, after step S1 and before step S2, the position error P err , observation speed V x and the observed acceleration A xTo determine whether there is a situation where there is no solution for the time delay, if so, an error is reported to the system terminal and the process is terminated; if there is no situation where there is no solution for the time delay, then determine the time delay time t dx The positive and negative signs of the time lag time are as follows: a positive time lag time indicates that the current actual position lags behind the command position, and a negative time lag time indicates that the current actual position leads the command position, that is, position tracking overshoot. Table 1 gives the basis for judging the sign of the time lag time.

[0047] Table 1 Delay time sign judgment

[0048]

[0049] In step S2, specifically, the time delay time is solved by the root-finding formula of a quadratic equation; for the two roots obtained by the solution, first determine whether the two roots meet the sign requirements based on the positive and negative signs of the predetermined time delay time; if there is no root that meets the requirements, an error is reported to the system terminal and the process is terminated; if there is a root that meets the requirements, it is used as the solved time delay time; if both roots meet the requirements, the root with the smaller absolute value is taken as the solved time delay time.

[0050] It is understandable that for the acceleration A x If the encoder feedback speed is directly differentiated by the control cycle to obtain the feedback speed, there will be a large disturbance, which will affect the estimation effect of the time lag time. Figure 2 The observer shown can effectively solve A x The computational disturbance problem can be solved, and the single-axis time delay time can be accurately estimated by calculation.

[0051] Furthermore, in step S4, when cross compensation is performed, the control cycle of the cross compensation is made consistent with the calculation cycle of the current loop regulator, and data transparent transmission between the axes is achieved by means of high-speed industrial Ethernet. The present invention combines the cross compensation control strategy to achieve coupling compensation between each axis and the virtual time-delay axis, coordinate the response performance of each axis, and realize multi-axis synchronous control.

[0052] Reference Figure 3 As shown in the figure, it is a block diagram of the multi-axis coordinated control strategy based on the virtual time-delay main axis, and the block diagrams of the z-axis and a-axis are omitted in the figure. Taking the x-axis in the figure as an example, P xr , P xf 、V xf and C xf They are x-axis position setting, position feedback, speed feedback and current feedback respectively; G px , G sx and G cx Respectively represent the position loop regulator, speed loop regulator and current loop regulator; G td For a cross-coupled regulator, a PI regulator can be selected; t dxThe observed x-axis lag time is calculated by averaging the lag times of each axis to obtain the expected lag time t d , through the cross-coupled regulator G td It is superimposed on the current loop regulator to adjust the operation of the permanent magnet synchronous motor PMSM and achieve cross compensation.

[0053] In summary, the multi-axis synchronous coordinated control method based on the virtual time-delay spindle provided by the present invention equates the response characteristics and operating status of each axis to the "time lag" time between the actual position and the command position, constructs a virtual time-delay spindle with the goal of consistent "time lag" of each axis, and then realizes the coupling compensation of each axis with the virtual time-delay axis, and finally coordinates the response performance of each axis to realize multi-axis synchronous control. Compared with the traditional solution, this solution has stronger universal applicability. It overcomes the limitation of the traditional virtual spindle solution that relies on inherent synchronization to construct the virtual axis; in actual application, each axis only needs to estimate the time lag time and coupling compensation value of the axis, and the complexity of the solution is independent of the number of system axes; the time lag time standardizes and normalizes the operating status of each axis, and the parameters of the compensation regulator do not need to be adjusted according to factors such as curve type and feed level.

[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A multi-axis synchronous coordinated control method based on a virtual time-delay spindle, characterized in that: The steps include: S1. For any axis in the multi-axis system, represented by the x-axis, an observer is established to obtain the current position error P of the x-axis. err , observation speed V x and the observed acceleration A x ; S2, the time lag time t corresponding to the x-axis at the current moment is obtained by the following formula dx : S3, for each axis in the multi-axis system, the method described in steps S1-S2 is used to obtain the time lag time corresponding to each axis at the current moment; S4, adopt the cross compensation control strategy, average the time delay time corresponding to each axis to obtain the desired time delay time t d , and is superimposed on the current loop regulator of the multi-axis system through the cross-coupling regulator to achieve cross compensation.

2. The multi-axis synchronous coordinated control method based on virtual time-delay spindle according to claim 1 is characterized in that: In step S1, when establishing the observer, the observer input includes the x-axis position given signal P xr The current signal C corresponding to the motor system feedback xf ; The position given signal P xr Accept observation speed V x The integral of is used as negative feedback to form the position error P err ; Position error P err After passing through the PID controller, the output acceleration signal a co , acceleration signal a co Considered to be caused by the resistance of the motor system; Current signal C xf Multiply by the motor torque coefficient K t , the electromagnetic torque T can be obtained em , and then divided by the system inertia J, we can get the acceleration a generated by the electromagnetic torque e ; Then the observed acceleration A x =a e +a co ; Observed acceleration A x After integration, we get the observed velocity V x And participate in speed loop regulation.

3. The multi-axis synchronous coordinated control method based on virtual time-delay spindle according to claim 2 is characterized in that: After step S1 and before step S2, the position error P err , observation speed V x and the observed acceleration A x To determine whether there is a situation where the time delay is unsolvable, if so, an error is reported to the system terminal and the process is terminated; If there is no time delay without a solution, then determine the time delay t dx The positive and negative signs of the time lag time are as follows: a positive time lag time indicates that the current actual position lags behind the command position, and a negative time lag time indicates that the current actual position is ahead of the command position, and then the process goes to step S2.

4. The multi-axis synchronous coordinated control method based on virtual time-delay spindle according to claim 3 is characterized in that: In step S2, specifically, the time lag time is solved by a root-finding formula of a quadratic equation; For the two roots obtained, first determine whether the two roots meet the sign requirement based on the sign of the predetermined time delay; If there is no root that meets the requirements, an error is reported to the system terminal and the process is terminated; If there is a root that meets the requirements, it is used as the solved time delay; If both roots meet the requirements, take the root with the smaller absolute value as the solved time lag.

5. The multi-axis synchronous coordinated control method based on virtual time-delay spindle according to claim 4 is characterized in that: In step S4, when cross compensation is performed, the control period of the cross compensation is made consistent with the calculation period of the current loop regulator, and data transparent transmission between the axes is achieved by means of high-speed industrial Ethernet.

Images