Integrators and the bird’s-eye view
Although specialization is often valued in one’s career and an early focus on a particular field is encouraged, for the general progress of society, so-called integrators, or “Jacks-of-all-trades,” are necessary, who can see the problem from the bird’s-eye view.
Mechatronics is a branch of technical sciences that connects mechanical engineering, electrical engineering, and computer engineering. Understanding these areas and their interdependencies is essential for the development of mechatronic systems. Examples can be seen all around us:
- Transportation vehicles: cars, trains, airplanes, ships
- Robotic systems: mobile robots, factory robots
- Various-purpose vehicles: cranes, forklifts, agricultural machinery, elevators
The interaction of different systems is clearly visible in the example of a car, which contains mechanical parts (wheels, engine, and transmission), electronic circuits (engine control unit, infotainment system, lights, and signaling), and software that manages electronic units. Since changes in one subsystem affect the others, connecting individual parts into the final product is not a simple task.
Control systems
A very important aspect of any mechatronic system is the control of the system itself. If a certain product does not perform the function we envisioned, is it even useful to us? This question is addressed by control theory, and every control system consists of three fundamental parts:
- gathering information about the current state of the system and its surroundings (sensors)
- determining the next step using rules of operation, i.e., control logic
- implementing changes on the real system (actuators)
Together, these parts work in a so-called closed loop using feedback, as shown in the image below.
Examples of control systems are not always obvious, but they are also very present:
- thermostat (temperature sensor – control logic – heater)
- cruise control in cars (speed measurement – control logic – motor control)
- Anti-lock Braking System (wheel slip detection – brake logic – brake control)
Depending on the purpose, the control logic can be very simple (e.g., just turning a heater on or off), but also quite complex (e.g., when managing the overall energy consumption in a building).
The PID controller
Imagine you want to control the temperature in a room, i.e., keep it at a certain set value. For this, you will need a thermometer, some control logic (thermostat), and an air conditioning unit. The subject of control theory is the development of control logic, which in this case can be based on the following simple rules:
- “The further we are from the set temperature, the more effort we need in order to reach it.” This means that the heater will produce more heat if the temperature difference is 10 degrees compared to a difference of 3 degrees.
- “If we have been away from the set temperature for a longer time, we need to work harder to reach it.” In case there has been a change in the system, e.g., opening a window, it’s possible that the first rule alone won’t be sufficient to achieve the goal, and there will be a constant offset. Then the second rule serves to eliminate errors that persist for some time.
- “If the deviation is decreasing rapidly, we need to slow down to prevent overshooting.” As we approach the target, there is a possibility of overshooting the set value due to system inertia (especially in thermal systems, which react relatively slowly). By gradually slowing down depending on the rate of approach to the target, it’s possible to avoid large oscillations.
Interestingly, this control logic requires almost no knowledge of the system (e.g., model, parameters, etc.), but is based solely on the deviation from the set value, i.e., the control error. Thanks to its simplicity, it is the most common form of controller in practice, known as the PID controller. The name comes from the aforementioned rules, where the first rule corresponds to the so-called Proportional control, the second rule to Integral control, and the third rule to Derivative control.
What about more complex systems?
It has been shown that PID controllers work very well in practice, are simple to implement, and easy to maintain. However, they are most commonly limited to systems with a single input and a single output (e.g., when using a single heater to control the temperature in one room). What if we would like to control larger systems with several inputs and outputs? Examples of such systems include various plants and buildings, but also modern vehicles, and even drones.
Of course, it is possible to use several PID controllers, each controlling a different part of the system. However, often we can do better than that! And how exactly would that work? Find out in the next article 🙂
About the author
Josip Kir Hromatko completed his undergraduate studies in Electrical Engineering (Automation track) at the Faculty of Electrical Engineering and Computing in Zagreb, and his graduate studies at Chalmers University of Technology in Gothenburg, Sweden. Since 2021, he has been a member of the Laboratory for mechatronics at FER, and he is currently employed as an assistant on the HRZZ project “PVDC – Predictive Vehicle Dynamics Control“. In late 2022, he enrolled in a doctoral program with a research focus on the same topic.