Servo controllers typically use a traditional cascade control method with a position loop embedded in a speed loop. In the era when this control method originated, current and speed control were implemented in hardware, while position control was implemented in software. One reason for its continued popularity is its simplicity. First, the speed loop is adjusted, then the position loop, while current control parameters are generally set automatically. The position controller usually consists of a simple proportional coefficient, while the speed controller includes a proportional coefficient and an integral element, as shown in Figure 1.
One drawback of this control method is the presence of a speed-proportional following error. Using feedforward control can reduce this following error, but at the cost of overshoot or increased settling time.
To overcome the aforementioned limitations and optimize servo performance in high-precision motion control applications, a unique algorithm called HD Control (HDC) employs a parallel control approach, where all branches operate at the same level and execute simultaneously within a single sampling period. Each branch includes a variable gain parameter that is automatically optimized to achieve high gain and high stability. Consequently, position errors and settling time are minimized, significantly outperforming other controllers.
Advantages of HD control
Minimize position error
Near-zero settling time
No overshoot at the end of the deceleration phase
No oscillations in steady state
Minimize vibration in steady state
Strong anti-interference ability
High path following accuracy
The algorithm mainly consists of two modules: a variable gain module to reduce the following error, and an adaptive feedforward module to reduce the tuning time. (See Figure 2.)
Variable gain (VG) control
During the HDC algorithm operation, the variable gains (VGd, VGp, VGiv, and VGi) are automatically calculated and dynamically modified. At the system variable level, each gain has its specific function, such as velocity error and position error. During motion, the variable gain may be ten times higher than when the motion is stopped. This ensures high accuracy in path following during motion and at lower speeds. In addition, during motion, the system rigidity is increased by more than three times, thereby ensuring very small following errors.
The four variable gains are balanced through a unique algorithm to ensure system stability. The branch containing the Kd parameter is similar to the velocity feedback loop and is used to reduce velocity error. The branch containing the Kp parameter is a proportional position feedback loop and is used to reduce position error. The branch containing the Ki parameter is the integral element of the position feedback loop and is used to reduce steady-state error.
The branch containing the Kiv parameter is unique to HD control, combining the effects of the Kp and ki branches. It produces more than twice the stiffness of Kp without oscillation. It is used to reduce following errors in both acceleration and steady-state phases. It can also be used to eliminate steady-state errors like the integral parameter (Ki), but with a response speed as fast as the proportional parameter (Kp). See Figure 3.
Adaptive feedforward
The adaptive feedforward module is used to achieve a very short settling time. Due to the superior performance of the Kiv and Ki branches, most of the feedback response (current command) resides in the integral stage. During motion, the adaptive feedforward module monitors the consistency between acceleration and motor torque and uses this relationship to handle the integral stage during deceleration.
At the end of the motion, the adaptive feedforward algorithm modifies the parameters in the integral stage according to the desired path acceleration, thereby achieving zero settling time, as shown in Figure 4.
Automatic adjustment
The HDC algorithm has been applied to the CDHD servo drive series developed and manufactured by Servotronix, as shown in Figure 5.
Parameter tuning is performed automatically through the CDHD user interface software ServoStudio™. However, automatic adjustment is often insufficient, and some applications require manual fine-tuning to optimize parameters.
Both automatic and manual adjustments are based on the same principle. In automatic adjustment, the quality of motion is measured and evaluated by the driver and software. In manual adjustment, the evaluation is done by the user. Regardless of the method, the servo control parameters are progressively modified to select the value that achieves optimal performance.
HDC parameter tuning is simple and intuitive, and its execution is very similar to traditional PID parameter tuning. Each variable gain is gradually increased until oscillation occurs, and then reduced by 10-20% to return to a safe range.
Applications of HD control
A Servotronix customer's gantry robot application required a sustained accuracy of 2-3 micrometers while operating at maximum speed. Using a CDHD servo drive with HDC algorithm, the application's maximum operating speed was increased from 120mm/s to 160mm/s while maintaining accuracy, resulting in a 33% increase in production efficiency.
In a comparative test with servo drives from other manufacturers, the CDHD drive achieved significantly higher accuracy and lower ripple when the device was running at a speed of 160 mm/s, as shown in Figure 6.
In summary, HD control has proven to be quite advantageous in applications requiring high path tracking accuracy and low settling time, such as CNC and cutting, conveyor following, pick-and-place operations, PCB assembly, soldering, painting, spraying and gluing.
