Research On Multi-Objective Path Planning And Track Control Of Unconstrained Mobile

Ball and Plate system is a typical nonlinear control system, scholars both at home and abroad all have certain research in Ball and Plate system.The ball above a flat is called the sports body without restraint because of its own movement is relying on flat movement. Ball and Plate system is a typical system of study nonlinear control and the motion planning problem.In the process of the study to ball and plate system about motion planning, we consider the flat upper part the camera as an external information perceptual system, we can collect real-time image to get the current environmental information, and then do reasearch to the motion planning problem, so it has many similarities to the traditional path planning problem for instance the robot path planning and vehicle navigation systems, but also because of its own mechanical structure characteristics, that is to say the ball does not has mechanical constraint, so they have different place. this is another research for ball and plate system, trajectory tracking control problem.This paper aiming at this kind of typical nonlinear system, studies multi-objective motion planning problem and trajectory tracking control problem. In the process of the research for multi-objective motion planning problem, many domestic and foreign scholars have adopted such as fuzzy neural network and intelligent algorithms, as well as in trajectory tracking control such as fuzzy algorithm.In the reasearch of the ball and plate system of laboratory,during the study of multi-objective motion planning,this paper does some work as follows:1.Environment modeling and ball position extraction.The multi-objective motion planning problem of ball and plate system in this paper is an offline path planning.Its essence is first planning out the right path, and then tracking. So the first problem is environmental problems. We must know the location of the obstancles,trap and ball position in the environment map, then identified to planning. On the basis of the existing research of the laboratory, during the reasearch of the image histograms, according to the characteristics of the experience,we find that the pixel of the trap is below 30, therefore may first find the trap, and by using the histogram which has three peak and two valley pixels to binary processing, wo can find the position of obstacles and the ball.At the same time we use the grid to describe the image, and in this process the trap grid will be marked - 1, the obstancle will be marked as 1, and the free grid position for 0, which make prepares for the next multi-objective motion planning.2. Multi-objective motion planning problem of unconstrained mobile establish and expression. In this paper, the path planning study of multi-objective problem is one of the most important question. For the goals of path length, risk and time of expression, laboratory work has no preliminary studies because the risk has no mathematical model and the time is considered an hypothesis. This paper gives different risk model to the obstacles and the trap, and introduces a metric function to measure time volume, such wo can give an exact mathematical expression for path length, risk and time. In order to make unified on three physical quantities, this paper also introduces multi-objective fuzzy optimization method to determine the final evaluation function,which is path search direction of metrics.At the same time,this paper gives the solution process of motion planning.The path according to the evaluation function will eventually be the smallest direction search until target.3.The application of Tracking Control Algorithm. According to the characteristics of multi-objective motion planning of the ball and plate system, combined with PID parameters on the characteristics of the control system. Fuzzy PID controller is used to calculate the expected rotation angle plate, and the difference between the actual angle of rotation for the inner ring after the input of a PD controller to calculate the number of pulses needed to turn the motor, thereby driving the motor rotation, so that the ball along the planning to track the path out until the end.4.Experimental results on multi-objective motion planning problems for ball and plate system. Using images collected through the environmental map obtained after process, the problem of multi-objective weights given according to the distribution of weight needed after the multi-objective motion planning by path searching, more in line with actual requirements and people subjective desire. We find that during experiments in different conditions, system planning out of the path is different. We did experiments in different environment, the same weight and the same environment and different weight to make a comparation.The final results show that the multi-objective motion planning method can find a different path to meet the different requirements of people, some may has a shorter path length, some may has a higher safety factor, while others may be the shortest time. In this case, depends on the value of people's desire to choose. Different weights will produce different paths.5. During the experimental studies, this paper has also established a competition evaluation index system, while conducting a preliminary study to the minimum energy path planning. The purpose of establish competition evaluation index system is for different experiments, we can give a standard of comparison to judge which experiment is better. Evaluation index system of competition based on the completion of the experiment is good or bad, for example, the size of path length, the length of time consumed, whether fall into the trap of tracking or run into obstacles and so on.Energy minimum path is also a path planning research focus, this paper analyzes the physical structure and force of the system to conduct a simple study.The energy integral of the ultimate source of energy consumption is used to obtain the formula and conduct a simple experimental study.In conclusion, this paper is to study the multi-objective motion planning problem in a complex environment of obstacles and traps of the ball and plate system. Different risk model of obstacles and traps are given for this complex environment,while introducing a measure of the size expressed as a function of time.Then multi-objective fuzzy optimization method is used to establish a multi-objective motion planning evaluation function of the system and gives a clear mathematical expression and solution process.Do some reasearch of the tracking controller and establish a competition evaluation index system as well as making a preliminary study to the minimum energy path and do a series of compare experiments in different conditions.