Institutional Repository of Vilnius University
Lazerio spindulio pozicionavimas naudojant PID valdiklį su Smith prediktorium /
Abstract
dc:descriptionThis is a Bachelor's thesis by a student of Vilnius University of the Physics faculty, Augustas Kunc. The goal of the thesis is to solve the problem of what should the angle of the drone be in order that the drone's position on an axis is equal to zero aka is centered? The Center for Physical and Technological Sciences is running a project whose main idea is to measure air pollution using two drones, two lasers and one reflector. So there will be one drone with a laser and four detectors, and another with a reflector that reflects that laser very well and back in the direction that it came. It will then be possible to send a beam of the second laser through the air and compare it with the returned beam to determine the air quality in that area. This paper will only focus on the positioning — the fact that these drones need to be facing each other as accurately as possible. At first, the drones will have to be at least roughly facing each other so that the outgoing beam reaches the other drone at least in a very small part. If this condition is met, it means that the outgoing beam will reach the reflector, which reflects the signal that enters it from any angle back in the same direction. The returning beam will hit the four detectors and saturate them. Saturation will have some value. By subtracting the saturation of one detector from the other, it will be possible to judge the position, for example, of the drone in the horizontal x axis, i.e. how much it is turning to one side or the other. So by changing the angle of rotation of the drone, one can observe how its position changes. The goal is to have the same saturation in both detectors and have the laser shining toward the center. In this paper, we will apply automatic control. All information of the system will be in two variables - input and output. In general it can be anything, e.g. the output is the temperature of the water in the pot, and the input is the gas knob regulating the flame that heats the pot of water. In our case, the rotation angle of the drone will be the input, and the received signal of the position of the drone, calculated from the satiation of the detectors, will be the output. We will apply a PID controller. That alone would be sufficient if there was no delay in the system. Effective delay compensation will be done using the Smith predictor. What a Smith predictor does can be explained simply like so. Let us take a scenario where we had a good PID controller for a system that did not have a delay, but then nothing about the system per se changed in any way, but there appeared a delay, that delay will be effectively compensated by the Smith predictor. The conclusions of the work are as follows. From the input and output signals, it is possible to recover the transfer function of the system by filtering the signals and applying the ordinary least squares method. Transfer function recovery is also possible for a linear system with a delay, i.e. when the input signal to the system is delayed. It is important to consider the possibility that there will be latency in the system. Because when restoring the transfer function as if there was no delay in the system, for such a system that actually does have a delay, the coefficients of the restored transfer function may be very inaccurate and poorly describe the system. We can see that when there is no delay in the system and a simple PID controller is used, we get good control of the system. However, when a delay occurs in such a system, the control of the PID controller becomes less stable, with more perturbations. Meanwhile, after the Smith predictor is applied to the system, its control becomes significantly more stable. Disturbances disappear and the whole control form resembles a system in which there is no delay at all. Although it is more difficult to implement the Smith predictor, it is possible to obtain stability characteristics as if there is no delay.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Kunc, Augustas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD192833569&prefLang=en_US