Report On Automatic Parking Assist System
Vehicle automation, autonomy and connectivity is a subject of mechatronics integrating many engineering disciplines including electrical, mechanical, control, computer engineering and technology. It is phenomenally changing the concept of automobile technological research and manufacturing. This project presents two major components of an automatic parking assist system (APAS). The mapping of the environment of the vehicle and detection of the existing accessible parking spot where the vehicle can park into is performed by APAS.
The two most vital tasks the APAS is to realize the design of feasible path geometry and the tracking of this reference in closed loop such that the velocity of the vehicle is generated by the driver and the controller determines the steering wheel angle which is realized by Steering controls of the vehicle. Here we present in the detailed analysis of the continuous curvature path planning problem and the time-scaling based tracking controller. The control strategy employs the minimum turning radius of the vehicle by means of distance ultrasonic sensors to determine the parking path. The programming strategy can simplify the analysis of the parking path; therefore, we do not need to apply any expensive sensors and complex mathematical calculation to determine the parking path. A vehicle could be parked with precession and safety in the parking space by following these routes.
The car-like robot uses Arduino microcontroller unit as the core for system integration and data computation. It controls peripheral hardware of the car body, ultrasonic range sensor. To avoid radial imprecision and to detect the obstacles along its path we will be using eight ultrasonic range sensors. Image processing does not give accuracy in an outdoor environment due to continuous changes in physical conditions of the surroundings. Therefore we will be using an ultrasonic range sensor to detect environments, and design a fuzzy logic controller with Analytic Hierarchy Process (AHP) and concepts of Artificial Neural Networks (ANN). Our motive is to reduce the time of parking correctly in the parking space, thus accelerating the parking efficiently.
Introduction
An important goal of automatic vehicle control is to improve safety and driver’s comfort. Parallel parking is considered as one of the most tedious way to park by most of the drivers. Drivers lack in judging the precise measurements in order to maneuver. Nonetheless, most of the challenging tasks can be performed through practice, common sense and training in following the rules which can be expressed in the form of “If…Then…” logic. These “If…Then…” logics are the primitive notions behind Fuzzy Logic control. Parking a vehicle requires the driver to understand over time from constant exposure from the environment. We can say that this process is similar to neural network wherein the mathematical models are formulated from the driver’s parking strategy. The APAS are generally implemented for low velocities parking motions. The APAS gather details about position of the obstacles and the vehicle surroundings first which is important to find an available parking slot and to complete a safe parking. We may classify automation in parking system according to the level of automation involved while maneuvering. In the case of high automation the APAS controls both the steering angle and the velocity of the car.
In the case of vehicles without automatic gear the low automation in APAS is the only available option. In this case the APAS controls the steering wheel and the driver controls the ABS and acceleration.
Nowadays, various APAS system with different level of automations are available in the market. The Volkswagen offers the driver with an option that also assists parking maneuvers. This solution uses ultrasonic sensors for low level of automation in parking maneuvers. Whereas, Valeo’s PARK4U claim that they can parks a vehicle in just a few seconds. Ultrasonic sensors are used scan the sides of the road to detect a suitable parking spot. The parking maneuver takes place in the usual way but in hands free mode. If the specification of the vehicle model is known, then the path planning method and the controller algorithm can be developed based on the equations of the mathematical models. Motions can be anticipated with deterministic or stochastic methods. Algorithms based on ANN and fuzzy logic can be applied to obtain an optimized path.
The controller in highly automatic vehicles may affect the performance of the car using two inputs. Whereas in one input, in the low automation where the car velocity is manually controlled hence the velocity during the execution of the parking might considerably differ from the one used in path planning. This problem is solved by using time-scaling. Our goal is to develop a parking assistance system which can execute in high automation in three different parking situations (perpendicular parking, parallel parking and diagonal parking).
Detail Design
The controller controls the steering angle by taking the inputs from the sensors as the distances and also is in charge of ABS. Once the system has anticipated its current configuration at each sampling point, the inputs can be processed and updated by the controllers directly. After the desired turning angle is obtained, the braking is generated by the use of NFC.
The conditions of the next time interval can estimate by evaluating the kinematics equations with mathematical models and AHP. The structured way is to generate the steering angle and the ABS control to perform the parking which is calculated by measuring distances covered by the all eight sensors against the obstacle present.
In the simulation APAS, the user gives the command of automatic parking. When the CLMR finds the appropriate parking spot, it keeps moving until it reaches the point where the it starts reversing. At this position, the environment data is retrieved by ultrasonic sensors are fed back, and then the membership function of NFC is adjusted by AHP before moving backward. The NFC changes the moving direction of CLMR, so the automatic parking is completed until the optimized parking position is achieved.
Conclusion
The goal was to give real-time solutions for autonomous systems. The project features sensor-based obstacle detection parking using NFC system. APAS has various controls to consider in maneuvering the vehicle: the position of CLMR with respect to the hindrances and the parking spot, the angle of the steering wheel, and the velocity of the CLMR. The controls have been trained through the fuzzy inference system then applied into the framework of ANN. The NFC is used to evaluate the actual sensor data with the help of AHP obtained from the environment to enable the CLMR for proper maneuvering.
