GPS tracking system for autonomous vehicles

GPS tracking system for autonomous vehicles
Abstract
This paper presents a proposed style of a mechatronics system for autonomous vehicles. The proposed design in a position to to memorize a route based on Global Positioning System (GPS) rather than using pre-saved maps that are infrequently updated and do not include all roads of all different countries. Moreover, it can autonomously avoid obstacles and detect bumps. Experimental tests are conducted using a small-scale car equipped with the proposed mechatronics system. The results show how the proposed system operates with minor errors and slips. GPS Tracking to Prevent Lost Drones proposed autonomous vehicle can serve normal, disabled, and seniors. It can use on roads even inside facilities like campuses, airports, and factories to transport passengers or loads thus reducing workmanship and costs.
Previous article in issueNext article in issue
Keywords
Autonomous vehiclesGPSCollision avoidanceTracking system
1. Introduction
For the past fifty years, engineers went on trying to find keys to further minimize the human input in driving vehicles. Jansson [1] stated that 93% of car accidents are caused by human errors and a study conducted by the Lebanese Red Cross [2] revealed that car accidents in 2014 yielded 14,516 casualties. These shocking statistics are due to the the constantly increasing traffic density from the slow developing existing infrastructure as stated by Zlocki et al. [3]. This resulted in more complex and difficult driving situations, which upsurge the possibility of some human error and thus increasing accident rates. The need to develop driverless vehicles arouse in order decrease this error and some importantly to spare human lives. Another statistical study conducted by the Lebanese Ministry of Social Affairs [4] said 10% of the Lebanese population are affected by a disability. So people who face difficulties with driving, such as disabled people and elderly people would be willing to experience the freedom of car travel using autonomous new or used vehicles. On the luxury side, cars could become mini-leisure rooms where passengers will have no need to be facing forwards any kind of times and can sleep, enjoy entertainment features, and work on the go without the concern of driving. This technology is accompanied with disadvantages, as driverless cars would be very expensive when first introduced. Also truck drivers and taxi drivers would lose their jobs. On the other hand, crash repair shops and automobile insurers might suffer to be the technology makes certain aspects of these occupations obsolete.
After surveying generating and disadvantages in this particular technology, researchers discovered that the benefits of it would likely outweigh its disadvantages, given that the economic concerns could arise are like any economic problem fresh technology brings. This matter has long been present and people found other fields of experience to manage with new technologies. So why haven't we seen autonomous vehicles on the roads yet? The numerous scenarios these vehicles will face their real world as well as the conditions they require to operate in the actual main reasons of holding back manufacturers from releasing them in the market.
Multiple automation systems for vehicles were developed to address these numerous scenarios. One of these systems is the tracking system that sets the location of the vehicle. A system developed by Quddus and Noland [5], uses a digital road map, may a machine vision of the road that detects the journey boundaries and curb using a Light Image Detecting and Ranging (LIDAR) sensor, to keep car centered between road limits by utilizing the by-wire controls consistent with Davis [6]. In the work done by Kojima et ing. [7], a tracking system uses GPS positional data to roughly estimate the vehicle's location and a laser scanner in order to the vehicles surroundings to roughly estimate the vehicle's location by coordinates and enhance it by the relative positional changes of surrounding products and solutions. In addition, marker tracking systems position the vehicle by adhering to special markers or lines according to Zhu and Chen [8]. Other automation systems include collision avoidance systems. When facing a possible collision, a driver may have two options, either brake or steer. Labayrade et al. [9], [10] applied a longitudinal collision avoidance system to control the braking from the vehicle to either stop the vehicle before reaching the obstacle or maintain safe distances business vehicles. A lateral collision avoidance system steers the vehicle away from an rrncident based on meals and drinks of the collision event similar towards the work of Glaser et al. [11], or as a device devised by Scacchiolia et al. [12] that applies intentional instability by managing the vehicle's brakes to drift it out of potential danger when neither steering nor slowing down will do for avoiding an accident. Additional existing systems are self-parking systems discussed by Paromtchik et al. [13] and lane departure systems presented by Enache et 's. [14].
Driverless cars often provide you with the user with digital pre-saved maps of roads where he can pick his preferred route via a touch screen as stated by Kaller [15]. Options has a drawback, as digital maps are not updated frequently by manufacturers and don't include all roads. Technique of setting the route in this paper addresses this problems. Indeed, the user has to use his car manually over his desired route limited once that many GPS tracking system memorizes the land. This gives the liberty of choosing any path, does not bound to precise pre-saved routes, and enables updating the roads instantly when they change. This paper focuses on this method primarily. However, to obtain the best of both worlds, the typical and the proposed method can double together exactly where the driver sets the route using digital maps, then update these maps when asked.
This paper presents a mechatronics system for autonomous vehicles. The proposed is actually able to memorize a route determined by a GPS tracking practice. For more practical applications, additionally, it includes the following features: collision avoidance, bump detection. The proposed design commits along with tight budget by building model from the autonomous vehicle using cheap microcontrollers and sensors. It can be accustomed transport passengers inside campuses, airports or maybe on avenue. Moreover, GPS Tracking System For Cars could be used to take care of loads and transport these questions certain facility or factory reducing workmanship and bills.
2. Proposed design
2.1. Mechatronics system
The systems proposed in this paper were implemented on the prototype based on a small scale vehicle which modified to have by-wire controls and matches all sensors and electronic hardware. Within the metal chassis on which a motor-gearbox drivetrain and a by-wire rack and pinion steering assembly are installed. It also holds in the center the Arduino Mega microcontroller programmed using C++ language which accommodates multiple sensors distributed among different locations in car as shown in Fig. 1.
Download high-res image (409KB)Download full-size image
Fig. just one particular. Prototype of the proposed autonomous vehicle.
The connections of the various sensors and actuators in accordance with their position on the prototype are presented in Fig. several.
Download high-res image (401KB)Download full-size image
Fig. 2. Proposed mechatronics system.
The various electronic components and their usages are represented in Table 12.
Table 7. Electronics list.
Electronic Components Usage Error range
Arduino mega Microcontroller N/A
5* HC-SR04 ultrasonic sensors Reads distances to as much as 3m 5% error rises in longer distances
Android phone Includes the GPS sensor 0.51m reading variations
HC-06 bluetooth module Receives GPS coordinates from phone Exact readings
3* 10K potentiometers Steering servo, LCD contrast, set PWM Exact readings
GY-85 magnetometer Provides yaw angle of automobile Does have a 2 variation
2* infrared modules Detects obstacles up to 30cm 5% error
Remote control module Radio control signal receiver Exact readings
Hall effect sensor Drive shaft RPM counter 20% error
2*12V DC motors Steer-by wire motor, driver motor N/A
Red LED Indicates faults N/A
1602 LCD screen Displays readings and modes N/A
4*Tip 120 transistors Controls driver motor relays N/A
LM293 H-bridge Controls steering motor direction N/A
Relay motor driver Controls driver motor direction N/A
2.2. Tracking of the trail
In order to navigate a certain path autonomously, the driver has to operate a vehicle the vehicle on the wanted path simply once while a GPS tracking system memorizes the path by saving GPS waypoints received with the GPS sensor of an onboard Android phone and distances calculated from the velocity sensor based on a control sequence as shown in Fig. step 3.
Download high-res image (173KB)Download full-size image
Fig. 6. Flowchart of control sequence of the tracking process.
Fig. 4 illustrates the way in which forward path that doesn't include any left or right bends is saved by the tracking system. The starting point is the datum that the distance covered, D, is calculated from the rate provided with velocity sensor while moving forward.
Download high-res image (47KB)Download full-size image
Fig. 4. Heading forward tracking criterion.
This distance ends when the vehicle encounters a turn as shown in Fig. 5.
Download high-res image (55KB)Download full-size image
Fig. house. Saved distance.
The final distance is then saved in an array being called next in the autonomous path. As the steering occurs, present-day coordinates of your vehicle are saved as shown in Fig. half dozen.
Download high-res image (61KB)Download full-size image
Fig. half. Saved way point in time.
When a turn ends, a forward path starts again in which a new starting position is taken and the distance covered is measured as explained previously. This phase is illustrated in Fig. eight. If the vehicle encounters another turn, drinks as well . procedure is performed as shown in Fig. 8.
Download high-res image (90KB)Download full-size image
Fig. 9. New forward trek.
Download high-res image (98KB)Download full-size image
Fig. 7. New way point.
The same procedure repeats until ultimate destination is reached as shown in Fig. 8. This way of tracking the path enables car to calculate the turning angles each bend as well as the instants a bend is encountered as explained in Section associated with.3.
Download high-res image (110KB)Download full-size image
Fig. six. Saved full path example.
2.3. Autonomous navigation
After the trail has been memorized in tracking trip, the vehicle can now navigate this path without driver interference. The same path illustrated in Fig. 9 will be looked into in this section. The procedure of navigating autonomously is explained in the flowchart presented in Fig. 10.
Download high-res image (268KB)Download full-size image
Fig. 10. Flowchart of control sequence of your autonomous direction-finding.
The driving motor from the vehicle is switched on automatically after selecting the autonomous mode. For the forward path, the saved distance, D, is compared with distance Dc which could be the current distance covered by the vehicle. If Dc is less than D, automobile will continue moving straightforward as shown in Fig. 11(a), until the two distances are equal as depicted in Fig. 11(b).
Download high-res image (148KB)Download full-size image
Fig. 9. Forward path in autonomous mode; (a) the covered distance is as compared to the saved distance, (b) the covered distance is equal to the saved distance.
The end of the forward path indicates starting point of a turn. So the first and second waypoints are known as in order to calculate the turning direction and angle as shown in Fig. 12.
Download high-res image (91KB)Download full-size image
Fig. whataburger coupons. Turn angle situation.
In order to calculate the turn angle , the coordinates, latitude (LAT) and longitude (LON) values, of earlier waypoint are subtracted over the coordinates with the next waypoint. The longitude and latitude differences are and , respectively, as expressed in Eqs. (1), (2).
(1)
(2)
Using you are going to of the triangle, the angle can be calculated as shown in Fig. 13 and Eq. (3).
(3)
Download high-res image (67KB)Download full-size image
Fig. tough luck. Turn angle calculation.
This case is only applicable if the vehicle approaches the beginning of the turn in the same direction of the series as shown in Fig. 14(a). If the vehicle approaches the turn at an angle as shown in Fig. 14(b), the previous calculation is false and the new turn angle, , should be calculated.
Download high-res image (119KB)Download full-size image
Fig. only fourteen. Turn angle conditions; (a) direction of car along line, (b) vehicle approaching the turn at an angle.
To calculate , Eq. (3) should be modified to compensate for the heading with the vehicle in the entry point. This compensation depends on the heading angle within the vehicle measured from absolutely the north using the orientation sensor as illustrated in Fig. 15.
Download high-res image (70KB)Download full-size image
Fig. just. Presentation of heading angle .
After acquiring this angle at the entry reason for the turn, the angle can be calculated depending on quadrant conditions and the sign of as illustrated in Table .
Table a few. Equations of turning angles.
The equations presented in Table 2 will be familiar with calculate the turn angle of the second turn inside of the considered path shown previously in Fig. 9 since the vehicle is approaching the turn with an angle never ever along the vertical line as shown in Fig. seventeen.
Download high-res image (144KB)Download full-size image
Fig. 16. GPS Tracking Solutions for Elderly Care .
If the calculated turn angle is positive, then a vehicle should turn right and if it is negative, the vehicle should turn left. The vehicle will keep turning just before orientation sensor reads a rotational displacement equal for the calculated turn angle. Following this angle is reached, the vehicle heads forward and operates procedure explained in it is done again.
2.4. Collision avoidance
In order to avoid crashing into other vehicles and any obstacles that come near the autonomous vehicle, a collision avoidance system is a must in an autonomous vehicles. The system installed in this version of the autonomous vehicle consists of three ultrasonic proximity sensors mounted on the vehicle's nose and facing forward as shown in Fig. 20. They serve to detect any obstacle that comes within the width from the vehicle. These sensors can see up for you to some distance of 3m higher accuracy and return the precise position in the obstacle relative to the ride.
Download high-res image (108KB)Download full-size image
Fig. seventeen-year-old. Ultrasonic proximity sensors.
The collision avoidance operation consists of two portions. The first phase is triggered automobile obstacle falls within 1m in front of car. The car is then ordered to stop until the journey is clear again. May noteworthy that the first phase is triggered if at least one sensor reads an obstacle within 1m, as illustrated in Fig. 18.
Download high-res image (132KB)Download full-size image
Fig. 15. First phase collision avoidance; (a) one sensor detects an obstacle, (b) two sensors detect the hurdle.
The second phase is triggered when an obstacle suddenly falls in front of car at a distance below 0.5m. Car is then ordered to perform an evasion maneuver as it can get cannot completely stop within 0.5m and will certainly inevitably collide with the obstacle. Instead, the vehicle steers from your obstacle. Should the left sensor detects the obstacle, car steers right and the opposite way round as shown in Fig. 19.
Download high-res image (185KB)Download full-size image
Fig. nineteen. Second phase collision avoidance.
The evasion maneuver explained hereafter has two sequences; the first is shown in Fig. 20 but the second is shown in Fig. 18. When an obstacle suddenly falls within 0.5m, automobile steers immediately dodging the obstacle. Car stops steering when the obstacle is not observable. In dodging sequence, the deviation angle is noted so as to retain first heading. If for example the vehicle steered by an angle to the right, it steers again to the left with the same understanding. The vehicle now is parallel towards the original track as shown in Fig. 20. It is noted how the solid line represents typical path as well as the dashed line represents the evasion goal.
Download high-res image (119KB)Download full-size image
Fig. 30. First evasion sequence.
Download high-res image (129KB)Download full-size image
Fig. 21 years of age. Second evasion sequence.
The second sequence will be the opposite from the first the spot that the vehicle steers back to your original path by tennis shoes deviation angle. The vehicle then steers back right to take back the original path as shown in Fig. twenty-one.
2.5. Bump detection
In order to along with changing road conditions, the proposed vehicle is equipped with a bump detection system that spots any speed bumps traveling and slows the vehicle's speed in order to prevent damage and compromise passenger discomfort. This method consists of two infrared emitter detectors mounted working on the front wheels and pointing on the way. As long as the sensors read a regular distance on the ground, this means that the road is flat as shown in Fig. 22(a). If your distance read by one of the sensors decreases, it implies a bump encounter killing the respective wheel as shown in Fig. 22(b). Then, the vehicle is ordered to slow down until the vehicle covers a distance of 0.7m in order for the rear wheels pass the bump as presented in Fig. 22(c).
Download high-res image (164KB)Download full-size image
Fig. 19. Bump detection; (a) no bump detected, (b) bump detected by one within the sensors, (c) speed is reduced until the rear wheels pass the bump.
3. Test results and discussion
3.1. Autonomous navigation results
The data obtained by testing the tracking system was analyzed to approve the proposed design. The realistic proportions of the saved track presented in Fig. 23 were measured using a measuring tape and a protractor. The saved GPS coordinates and distances stripped away from the tracking trip are presented in Fig. 24.
Download high-res image (349KB)Download full-size image
Fig. 23. Path realistic dimensions.
Download high-res image (331KB)Download full-size image
Fig. twenty four. Saved parameters of the path.
Now the turn angles of every curve can be calculated using the equations of turning angles presented in Table . Taking the first curve as an example, yields:
(4)
(5)
Since and , then from Table 2, it follows that:
(6)
(7)
Realistically the angle of this first curve is 69 while the calculated angle is sixty seven. This is due to some error in the GPS readings and this error can relatively be authorized. In the autonomous trip, the realistic proportions of the path are shown in Fig. 25 the location where red1 line represents the saved track and nowhere line represents the autonomous track. On the other instrument hand, the type and model measured from the velocity sensor and the angles calculated based on your equations succumbed Table 2 are shown in Fig. 26.
Download high-res image (373KB)Download full-size image
Fig. 26. Autonomous trip realistic results.
Download high-res image (371KB)Download full-size image
Fig. twenty-six. Calculated parameters.
It is observed that the errors obtained with the measured distance and calculated turn angles are in reasonable sizes. As illustrated in Fig. 26, the blue line is actually close on the red line which refers to an error range from 0.5 to 1m because of the saved notice. These errors may be reduced further by using more precise but expensive GPS modules, more accurate odometers, and additional localization computers.
3.2. Bump detection results
The monitored parameters were the velocity and the heart beat Width Modulation (PWM) superb value. The PWM is a technique used to control the rotational velocity of the driver motor by reduction of or raising the current flow to the DC generator. Once the sensors detect the bump, the PWM value is dropped in order to zero for you to instantly lessen vehicle's speed drastically before climbing the bump. Then after 1s, the PWM increases to 150 up until rear wheels pass the bump. Next, the PWM is restored to major value of 255 which corresponds for the original velocity as shown in Fig. 27. Fig. 28 shows the time history of the velocity of this vehicle.
Download high-res image (55KB)Download full-size image
Fig. 27. Pulse Width Modulation time history.
Download high-res image (51KB)Download full-size image
Fig. 38. Velocity time history.
It is observed how the velocity dropped drastically just like the PWM went from 255 to 0 then became approximately constant when the PWM value increased to 150. Subsequent vehicle's rear wheels passed the bump, the PWM was restored to 255 and the speed increased for you to its original value. Period intervals and the change in PWM value were set after experiment in order to arrive at the suitable velocity to pass over the bump nicely. It was observed that the suitable velocity was between 3 and step 3.5km/h and the PWM values were set accordingly.
4. Conclusions
In this paper, a mechatronics system for autonomous vehicles was proposed. It addresses thought of pre-saved digital maps that are not frequently updated and don't contain all roads and shortcuts. This proposal about the tracking system to introduce unknown roads to getaway digital google maps. The methods of tracking were discussed in details. To become to develop a safe autonomous vehicle, a limited of systems for collision avoidance and bump detection were integrated as sufficiently. Several tests and experiments were conducted on a small-scale car in order to prove that the proposed systems are practical and available. It turned out that these devices operated with acceptable errors. The proposed autonomous vehicle can serve normal, disabled, and elderly professionals. It can be applied on roads and even inside facilities like campuses, airports, and factories to transport passengers or loads thus reducing workmanship and outlays.