Elmer R. Magsino

Stabilizing quadrotor altitude and attitude through speed and torque control of BLDC motors

This research presents a control structure that utilizes the speed and torque control of the brushless DC motors (BLDC) within the attitude altitude loops in order to attain flight stability and maneuverability of the quadrotor Quadrotor platform. Classical controllers were used drive each BLDC motor to compensate for the errors in altitude and attitude of the quadrotor during flight. The control system was implemented on a dsPIC33 microcontroller that used various feedback and instrumentation sensors. A complementary filter was used to incorporate the data of the accelerometer and gyrometer to arrive at usable attitude estimates. The experimental results revealed that when using the proposed control structure, the attitude varied by less than five degrees while the altitude deviates by 50 cm. This was attributed to the fact that the PID controllers were able to compensate for the electronic speed controller (ESC) and motor mismatches. During one experiment, Quadrotor platform was able to reach six meters within five seconds.

A Blind Source Separation of instantaneous acoustic mixtures using Natural Gradient Method

A variety of applications concerning communication signal processing involves recovering unobserved signals or “sources” from several observed mixtures, and “cocktail party effect” is a good paradigm related to this process. Given a set of linearly superimposed acoustic signals without knowledge about the sources makes Blind Source Separation (BSS) a very suitable scheme. A more popular approach of BSS, Independent Component Analysis, has been exploited which basically senses the statistical independence of the source signal estimates to achieve separation. A set of interfering signals present in a typical acoustic environment has been instantaneously combined with a pre-determined mixing matrix. A great weight has been given on an excellent rendition of the Infomax technique of Independent Component Analysis (ICA), called the Natural Gradient Method, to employ a cost function that would yield an optimized de-mixing matrix, producing fairly estimated source signals. By varying the learning rate and the score function, a robust performance of the Natural Gradient has been exhibited, maximizing the separation quality, stability and convergence speed.

Dynamic analysis of a two-input Zeta converter topology for modular hybrid PV-wind microgrid system

This paper presents the dynamic modeling of a two-input Zeta converter topology using State-space averaging (SSA) technique. The modeling leads to a small-signal linear dynamic model of the multiple-input converter (MIC), from which the transfer functions used for feedback control design can be determined. To verify the accuracy of the obtained model, a two-input Zeta converter with inputs from PV array and wind turbine generation system (WTGS) is designed which can produce an output voltage of 350 Vdc and an output power of 1kW. A controller for this is designed based from the transfer functions derived from the dynamic analysis conducted. Simulation tests of the design are performed using blocks and component of Matlab/Simulink from its SimPower Systems Toolbox. Results are presented to verify the accuracy of the obtained model.

Elementary and simplified control of digital parallel buck converters

An elementary and simplified control of four digital parallel-connected buck converters was simulated in MATLAB/Simulink. These parallel converters, having the same values of inductances and capacitances but different parasitic resistances, were observed under various conditions of current sharing load and on the total number of active converters. The voltage and current errors of each converter were sampled and fed into a discrete controller to regulate its output voltage. An elementary and simplified control was used to determine which converters will turn ON/OFF during a given current load. Between active converters sharing the load, the average current method was used. In this paper we conclude that an elementary and simplified control of parallel digital converters is feasible and achievable leading to a possibility of a much more enhanced and improved control of digital parallel converters. Simulation-demonstrated results are presented including the static and dynamic loading responses, current sharing error of the parallel system under different conditions and the effective number of bits for each quantizer block.

A rapid screening algorithm using a quadrotor for crack detection on bridges

This research study employs the use of a quadrotor for improving the efficiency and safety of civil engineers in the inspection of concrete bridge building columns for probable cracks. At an initial scanning height, the quadrotor hovers vertically at constant height increments and captures the structure image. It communicates, via Wi-Fi, to an android device running the rapid screening algorithm for crack detection. Out of 23 collected images from a bridge structure, the average accuracy achieved was 95.65% with a precision of 91.67% and a recall of 83.33%.

Simulation of search-and-rescue and target surrounding algorithm techniques using Kilobots

This research study presents the simulation results, from the V-REP software, of using Kilobots in executing collective behaviors usually done in military operations. A given base configuration of 10 Kilobots are used in simulating both the target-surrounding and search-and-rescue behaviors. In each algorithm, a Kilobot performs three basic roles: (1) searcher/walker, (2) target and (3) base. In the target-surround behavior, a successful run is considered when five Kilobots have already encircled its target. In the search-and-rescue algorithm, three different techniques are presented, namely: (1) square formation, (2) dispersal and (3) sweep. Simulation results provide promising results for the hardware implementation of these proposed swarm behavior techniques though it will be seen the limitations of the V-REP software.

A redundant flight recovery system implementation during an octocopter failure

This research presents a redundant flight recovery system during a motor failure in an octocopter platform. This proposed recovery system exploits the property that the octocopter can be derived by combining two quadrotor systems. During motor failure, the octocopter switches to a stable quadrotor flight configuration while immediately landing the octocopter to prevent further damage to installed sensor and equipment and the octocopter platform. The redundant system utilizes PID and fuzzy logic controllers to stabilize altitude and attitude respectively.

Visual surveying control of an autonomous underwater vehicle

This paper implements a visual surveying control in providing motion control to an autonomous underwater vehicle (AUV). The AUV, which has a dimension of 35.5 cm by 16.5 cm, has five motors that are used to provide the necessary motions. Three motors are used for vertical movements while the remaining two are used for the AUVs horizontal movements. A programmable embedded vision sensor mounted inside the AUV processes and analyzes acquired information from its camera and then generates specific commands to be executed by the AUV. Visual surveying for motion control is achieved by recognizing color and tracking objects of a predefined shape. Color recognition aids the AUV to identify a specific color while object tracking makes the submarine to automatically move towards an object in response to its recognized color.

Equalizing Converter Utilization Using a Fuzzy Decision Mechanism based on Input Voltage Perturbation and Time of Usage

Abstract — In this paper, fuzzy logic was used to provide the information on the probability of turning ON of four parallel converters using the converter utilization parameter and the input voltage perturbation. Each converter is designed having equal values of inductances and capacitances but with different parasitic resistances and was modelled using MATLAB/Simulink environment. Based on the output probabilities, the appropriate converter/s to turn ON will be turned ON by an enable circuit written in MATLAB code resulting to a well-balanced utilization of all converters. The specified input voltage of each converter is set to be 20 volts with different variances. Since each converter utilizes different supply voltages, the supply voltage is fed to a Gaussian noise realized through a random generator block to provide the variance around 20 volts. The voltage and current errors of each converter were sampled and fed into a discrete controller to regulate its output voltage. Between active converters sharing the load, the average current sharing/ democratic sharing method was used. In this paper, it was found that the fuzzy logic controller can be used to control parallel digital converters to properly balance converter utilization. Simulation results are presented including the static and dynamic loading responses.

Offline and Simplified Industrial Automation Processes Employing Robotic Manipulators

One of the predominant applications of the theories and principles of robotics is in industrial automation particularly in automotive manufacturing. This paper explores some mini-projects related to industrial automation that will help students and even robotics enthusiasts understand the kinematics of each manipulator. These projects can be as simple as following a straight line to as complex as adding machine vision and online intelligence to the robotics manipulator.