Closed-loop control of the secret: the use of feedforward control system to make adjustment easier

Closed-loop control system has the advantage of: through the motion controller to make the advantages of hydraulic energy to play to meet the modern institutions to control more precise requirements. The control links containing the proportional, integral, derivative (P, I and D) gain parameters have become the standard function of the motion controller and additional control parameters have been added to optimize the control cnc machining center algorithm. For example, a feedforward gain is added to satisfy the increased dynamic system response while reducing position and velocity errors. This results in a higher machine control performance and a longer machine life due to smooth operation.
The motion controller uses a combination of proportional, integral and derivative gains to generate control signals to reduce errors in the target position and the actual position. The proportional gain (P) is simply multiplied by the instantaneous error between the target position and the actual position so as to proportionally act on the control signal at the next instant. The larger the error, the greater the resulting control signal.

Structural Steel Fabrication

The integral gain (I) is multiplied by the sum of the position errors over a period of time to produce an integral effect on the output. Even if the error is small at any time, the sum of the errors will eventually increase to a point where the error is reduced.
The derivative gain (D) is multiplied by the error of the target and actual speed [WHM1]. The effect of the derivative gain on the output control signal is proportional to the divergence Hydraulic notching machine and convergence rate of the target and actual position deviation. For simplicity, we ignore the differential gain by using a system with sufficient damping factor.
One limitation of using only PID control is that the proportional link requires an error to produce the control output, and the integral link requires error and time. The output signal used to control the electro-hydraulic valve is always related to the error between the target position and the actual position of the load. In many cases, if only the proportional gain is used, the error needs to be large enough to produce the desired control signal. Adding an integral link will be through the accumulation of errors, so that the output of the control signal increases, but the integral part of the output signal increases need time.
In general, point-to-point operation occurs very quickly, because the integrator does not have enough time for error accumulation, so here rarely used integrator. Even if the integrator accumulates the error, it is likely to cause overshoot of the position control when the error is reduced.
When the error signal between the target position and the actual position changes, the integrator error accumulation decreases. This phenomenon occurs only when the actual position is overshooted by a certain amount of the target position, which is usually not desired in the motion control system. The motion controller can reduce the number of components in the front feedback control loop, thereby reducing the frequency of operation of the integrator.
The front feedback uses the motion controller’s target or action curve generator information. In general, the control principle of a high-performance controller is that the motion controller generates the target motion curve and then controls the actual motion following the target motion curve by the control loop (just like a donkey in front of a donkey to guide it forward).
The target action is refreshed after a period of time and may be refreshed every millisecond. Before each PID controller is refreshed, the target generator calculates parameters such as the position angle iron machine, velocity, and acceleration expected by the action. Because the motion controller “knows” the target speed and acceleration, it does not need to wait for the PID control to respond to the target position and the actual position of the error, can be directly output to meet the speed and acceleration requirements of the control signal.
The strength of the output signal is determined by the front feedback gain, which is a predicted parameter. The PID gain is multiplied by the feedback error. Instead of the feedback gain, the pre-feedback gain is the predicted gain. Its parameters are multiplied by the target speed and acceleration, respectively, and the sum is added to produce the control signal. As shown in Figure 1. The control of the output current by the front feedback is based on a simple formula,
Front feedback output composition = Kv × target speed + Ka × target acceleration
Among them, Kv expresses the speed before feedback;
Ka represents acceleration before feedback.
It is worth noting that motion control can be achieved in different ways. As shown in Figure 1, many factors determine the size of the control current. The output current can be obtained from the negative feedback of the PID control, either from the positive feedback calculation or from a combination of the two. However, the control current obtained under the same conditions can reach the same speed. The system does not care whether the control current is derived from the previous feedback or in the PID control. We can observe the above phenomena in Figures 2a, 2b, 2c.
If the system does not have high-speed motion, why not just increase the PID gain? Because in order to ensure the stability of the system, for the calculation of the feedback gain can not be too large. Normally, the negative feedback gain is increased to achieve the requirement of reducing the error without causing oscillation and instability. The advantage of obtaining the control signal from the previous feedback is that the pre-feedback does not generate the control signal by means of the error signal, as PID control does.
The key to designing a stable and easy-to-debug system is to use as much feedback as possible to obtain the control signal and to minimize the use of PID control to obtain the control signal. So that the error of the result will be minimized.
The designer uses the PID unit to compensate for environmental angle bending machine factors such as temperature and humidity in the system and the non-linear part of the system response that changes over time (eg, changes in the system load) that can not be evaluated in the front feedback .
Figure 2a. Actual and ideal motion profiles obtained with the RMCTools software. Only proportional gain is used here. The actual speed and position differ greatly from the desired speed and position. The area between the actual position curve (red) and the target position curve (cyan) is the error necessary to produce the proportional gain. This error can be eliminated by increasing the feed forward link.
Figure 2b. Pre-velocity feedback is added to the system of Figure 2a. Due to the constant velocity, the position error (the area between the red and blue curves) has been much reduced, but there is still a position error (see the part of the oval curve) during acceleration and deceleration.
Figure 2c. The motion profile is obtained in the same system as in Figures 2a and 2b with simultaneous feedback of velocity and acceleration. The actual position curve is in good agreement with the ideal curve. This is an optimal regulation system.
The action of pre – speed feedback system
The motion controller uses the pre-speed feedback gain to calculate the control current that actuates the actuator at a given speed.

structural fabrication machine

By controlling the servo valve, the system makes one end of the hydraulic cylinder through the hydraulic oil, one end of the hydraulic oil, thus driving the hydraulic cylinder action. In order to move the cylinder at a given speed, the system needs to control the flow of the hydraulic fluid and ensure that the net force acting on the piston is equal to the sum of the traction of the load and the friction of the hydraulic cylinder.
We can empirically estimate the control current intensity at each speed. This is done automatically by the speed front feedback unit. As shown in Fig. 2b, the position accuracy of the system is higher after introducing the pre-velocity feedback gain.
For example, we give the valve a control signal of 10% of the maximum intensity of the control signal and measure the speed of the actuator. If the actuator travels at 1-foot-per-second at this point, we can estimate that the actuator is running at about 3 feet per second when the control signal is 30% of the maximum signal. In other words, the open-loop gain of the system is such that the system moves at 1% per second at 10% control current.
The pre-feedback gain should be set to the opposite of the open-loop gain of the system, so the front feedback of the system corresponds to 10% of the control current per 1-foot-per-second. The motion controller calculates the control current using the pre-feedback after a new target speed is given.
Speed ​​before the feedback adjustment
As mentioned above, we often rely on observing the system open-loop action to determine the gain before the speed of feedback. However, in some of the implementation of components and applications, we need to do more work.
For example, a single rod piston cylinder must be in the direction of extension and retraction direction, respectively, debugging. Because the piston rod on both sides of the role of different areas cnc drill machine in the system extension and retraction stage need to set a different PID control and the former feedback gain. By calculating the speed ratio of the protrusion and retraction, the gain ratio at the time of extension and retraction can be determined.
This means that we can determine the pre-feedback gain ratio by first inputting a control current of + 10% to the system and then inputting a control current of -10% to observe the velocity of the piston rod in different directions. Even if the actuator is symmetric, the system still has nonlinear links. Such as hanging heavy objects, the gravity will cause the actuator to move down and hinder the actuator upward action.
Ideally, the system can fully use the pre-feedback gain work. But in fact, the load will change and there are non-linear part of the system, the motion controller through the PID gain aided before the feedback function. If the system load is greater, requires a 32% control current instead of the previous 30% of the control current work. The PID element in the control loop can provide a control current gain of 2%. The pre-speed feedback provides 30% of the control current to provide the main power.

No Comments

rssComments RSS   transmitTrackBack Identifier URI

No comments. Be the first.

addLeave a comment