Figure 2: Concept map for a self-learning motion control system.

© Analog Devices

To implement this level of self-aware drive technology, we need to create a concept for the control system. Figure 2 shows the four key elements required for a successful implementation:

Element I - Goal or task: A clear goal or task must be defined for the system to accomplish. In our example, this means: "Move the beer mug from point A to point B in the best possible way so that no beer is spilled."

Element II - Desired system behavior: Once this goal is established, the next level of self-aware control initiates the desired movement behavior. For our beer mug example, this would be: "Use a linear motion to move the beer mug and automatically adjust the motion to compensate for the fluctuating weight and size of the beer mug within the required safety limits of the mechanical system."

Figure 3: Monitoring and automatic tuning of torque, flux, current, speed and position control loops.

© Analog Devices

Once the target and the desired system behavior have been defined, the adaptive control system ensures dynamic convergence between the kinematics of the core system and the associated mechanical system by automatically tuning the drive and the integrated mechanical system to achieve maximum operating performance in the respective working point (Fig. 2).

Element III - Core drive system: The heart of self-aware drive technology is its kinematics. The challenge is to observe, learn from and monitor the performance of the motor and drive system. To create a working model of the drive system, an intelligent observer must be implemented to gain a basic understanding of the motion parameters and the physical limits of the drive system. This is achieved through field-oriented control (FOC) with dedicated position sensors or a sensorless FOC approach to learn how the motor is loaded and reacts in its operating environment. By monitoring and automatically tuning the control parameter values from the control loops for torque, magnetic flux, current, speed and position of the motor, we can optimize the response of the drive system. Once these datagrams are collected and fed to the intelligent observer, the implemented optimization algorithm ensures that the motion control parameters are calculated and the underlying control algorithm converges to an optimal set of motion parameters (Figure 3). Now that an intermediate motion model has been created to model and optimize the motion of the drive system, the next stage of self-aware drive control can be implemented by introducing an adaptive control engine.

Figure 4: Adaptive control model.

© Analog Devices

Element IV - Adaptive control: Building on the kinematics and FOC autotuning capability of our system, we now focus on the next stage of implementing self-aware control, the adaptive control engine. This next level of intelligent drive focuses on communicating the desired system behavior to the adaptive control engine (Figure 4). This system behavior is provided by a production worker, the plant supervisor or by an AI productivity algorithm that collects plant data using its network of smart sensors. Once the desired behavior is passed to the adaptive control engine, the self-aware controller begins to dynamically reconfigure the operating parameters of the drive system to achieve the desired system behavior. Some examples of these desired behaviors are the requirement to increase production throughput or to extend the life of the motor by operating at safe operating points. While the motion controller automatically adjusts its parameters to achieve the new required performance level, the adaptive controller continuously monitors the closed loop to maintain the desired performance level. This state is maintained even if the drive system is subject to changes due to wear of the mechanical systems or if the operating conditions of the motor change. The system has now reached the highest level of self-aware drive control.

Practical example of beer transportation

Figure 5: Example of a self-learning motion control system in action (variable weight load).

© Analog Devices

The technology of a self-aware control system is currently being optimized for industrial use. One can imagine an entire factory based on devices with self-aware motors and intelligent sensors. This plant would have the ability to self-correct potential equipment failures, automatically adjust production processes to maximize productivity and extend equipment life throughout the manufacturing process.

The author: Jeff DeAngelis is Vice President of Industrial Communications and Motion Control, Industrial and Healthcare Division at Analog Devices.

© Analog Devices

The best way to illustrate this concept is with a real-life example (Figure 5). Let us now look at an example of the implementation of a self-aware control system: The aim of this task is to bring a glass of beer from the bartender (point A) to a customer at the other end of the bar (point B) in the fastest possible time without spilling a drop of beer. The equipment system in this case is a cup holder with a built-in weight sensor that detects the weight of the different sized beer mugs and moves them along the length of the bar using a linear motor. A self-aware drive system is beneficial to get the beer to the customer as quickly as possible, but it can also automatically adjust its speed and power when the customer places the empty or partially empty beer mug in the cup holder to return the beer mug to the bartender or have it refilled. This system also offers the possibility to adjust its behavior when the bartender uses glasses of different sizes for different types of drinks.