35results about How to "Efficient path planning" patented technology

The invention discloses an unmanned vehicle dynamic path programming method based on environment uncertainty, comprising steps of establishing a vehicle kinetics model, establishing a dynamic environment model and necessary conditions for anew programming a path, obtaining a motion state original value of the unmanned vehicle, a vehicle motion state original target value and a vehicle motion state candidate target value, generating a candidate path, obtaining an optimal path through selection based on a safety index and a rapidity index, and reprogramming the optimal path of the unmanned vehicle when the unmanned vehicle motion environment satisfies the condition where the new path can be programmed. Compared with the prior art, the invention can not only satisfy the safety requirement for safety driving, but guarantees the driving efficiency under the constraint of the vehicle model. The invention realizes the coordination optimization of the performance index through distribution of various weights, realizes the real-time planning under the condition where a plurality of dynamic obstacles exist, and effectively improves the safety of the unmanned vehicle driving.

The invention provides an unmanned vessel ant colony consumption optimal global path planning method based on an electronic nautical chart. The method comprises the steps of 1, processing the electronic nautical chart, wherein after the electronic nautical chart is converted into the form of a file capable of being processed, a feasible area and an obstacle area of a map are determined, a visibility graph method is utilized to find out a set of feasible paths among an initial point, a target point and the vertex of an obstacle, it is guaranteed that the paths do not pass through an unfeasiblearea, and the sailing safety of the unmanned surface vessel is guaranteed; 2, building a motion math model, wherein the motion model of an unmanned vessel is built, and thus the thrust and motion characteristics of the unmanned vessel are grasped; 3, obtaining unmanned vessel data, wherein when the unmanned vessel sails on the water surface, information data like the heading, the vehicle body position and vertical acceleration of the vehicle is obtained; 4, utilizing an ant colony algorithm to search for a track meeting requirements; 5, building an unmanned surface vessel energy consumption model under ocean current disturbance, wherein with the optimal energy consumption being the premise, the ant colony algorithm is improved to obtain the path satisfying safe sailing conditions and having the smallest energy consumption. By comparing the paths under two preconditions, the effectiveness of the designed algorithm is proved.

The invention discloses a driverless car local path planning method based on model prediction. The method comprises the steps: 1, establishing a driverless car kinematics model; 2, in the vehicle pathplanning process, achieving the approximation of a driving environment formed by a vehicle and an obstacle through multiple circles; 3, fitting a lane line and a road edge curve; 4, constructing a target function J, wherein a first item and a second item of the target function represent the deviation between the predicted value and the reference position of the center line of the main lane, the third item represents the deviation between the predicted output course angle and the main lane midline course angle so as to ensure that the vehicle course angle is the same as the road course angle and the front wheel turning angle does not change suddenly; 4, ensuring that the vehicle runs according to the reference speed; 5, in order to avoid obstacles existing on the current lane, avoiding thecollision. Compared with the prior art, the method guarantees the quick and efficient path planning, and facilitates the calculation of the shortest distance between the main vehicle and the surrounding vehicles and obstacles.

An embodiment of the invention discloses a vehicle access path planning method in an intelligent parking garage. A modeling method of an intelligent parking garage environment is presented, the AGV path planning problem is realized by combining the morphology dilation operation on the basis of modeling, and problems of dead locking, collision conflict and others during the path planning process are solved on the basis of a dynamic time window method. The method solves the environment modeling problem of the intelligent parking garage, and a map with similar environment as the parking garage can be manually generated instead of being generated by a sensor and the map can be used for the AGV work. The embodiment of the invention further provides a vehicle access path planning device in the intelligent parking garage, and a storage medium. By adopting the vehicle access path planning device in the intelligent parking garage and the storage medium, the vehicle access and parking tasks canbe smoothly completed within a short time on the basis of effectively avoiding the path resource competition and conflict. The method and the device have the advantages of high automation degree of vehicle access, high performance cost and high safety reliability, and the like.

The invention belongs to the field of emergency path planning, and particularly relates to a fire escape path planning method and system based on an extended A* algorithm. The method comprises the steps of: collecting building floor data, modelling the environment of each floor, and obtaining an N*N-dimensional map of each floor of a building, the map comprising a free grid for identifying a passing area, an obstacle grid for representing an obstacle area and a fire behavior grid for representing a fire area; and obtaining an escape path of an escape person on the free grid by utilizing an A*algorithm, and when a plurality of minimum cost values appear at a certain point of the map, taking each point with the minimum cost value as a next point for expansion so as to obtain a branch with the minimum comprehensive cost value, and taking the branch as an optimal escape path. According to the method, the shortest escape path can be obtained conveniently to achieve the purpose of correct and timely personnel evacuation, the casualty degree in a fire can be greatly reduced in combination with intelligent escape guidance of a path planning algorithm, and high practicability is achieved.

The invention discloses a path planning method integrating a dense convolution network and a competitive architecture. The method includes the following steps that: a mobile robot samples mini-batch pieces of conversion information from an experience playback memory, and selects one of two fusion path planning networks as an online network and the other as a target network according to preset rules; the maximum value of a predicted target action value function is obtained through a predicted online action value function Q(s, a; w) and a corresponding greedy action; a loss function at a currenttime step is calculated according to the maximum value of the predicted target action value function and the predicted online action value function; and an online network weight w is updated with a stochastic gradient descent method according to the loss function. According to the path planning method of the invention, the dense convolution network and the competitive architecture form a more lightweight fusion path planning network, and therefore, model parameters are simplified, training costs are reduced, planning time is shortened, and the requirements of high-speed path planning can be satisfied to a certain extent.

The invention discloses a mobile robot path planning method based on a network tearing method. A connecting line from a starting point to a target point is used as a basic driving route. An obstacle network model is established by using a network tearing method and sub networks are torn; mathematical fitting is performed and minimum weights of a path and time of a mobile robot are calculated respectively; an optimal path plan of arriving at a next obstacle and going around the next obstacle by the mobile robot is obtained; and the basic route is adjusted. According to the invention, the mobilerobot path planning method is simple and clear and has high implementability. The mobile path can be corrected effectively based on the obstacle and the path can be adjusted timely and flexibly basedon newly added obstacle, so that the optimal path plan is realized. When the mobile robot moves, the collision with the obstacle is avoided effectively, so that the service life of the mobile robot is prolonged.

The invention discloses a mobile robot path planning method. An EKF-SLAM algorithm optimized based on a Levenberg-Marquardt method, mobile robot global path planning based on a potential field ant colony algorithm, and an optimal path searching method integrating mobile robot global path planning and local path planning are provided; and after a robot simulation experiment, autonomous navigation of the indoor mobile robot is completed. Compared with a traditional method, the mobile robot path planning method has the following advantages: the stability and the positioning precision of the extended Kalman algorithm can be improved. And the effectiveness after improvement is verified in combination with a simulation experiment, so that a position prediction result is more accurate, and a better path plan is obtained.

The invention discloses a method and system for managing the real-time state of a passing road section of an indoor map, and the method comprises the steps: dividing all channels in a parking lot into a plurality of continuous road sections according to 5-7 meters, marking the road sections and intersections respectively, planning a preset route for entering or leaving vehicles according to the channels and parking spaces in the parking lot, acquiring a road section sequence and an intersection sequence of a preset route; determining an impassable section and an unavailable parking space, specifically, presetting a monitoring duration T and a fault threshold value; acquiring a real-time route of the vehicle, generating a to-be-confirmed section, judging whether the to-be-confirmed section is an initial fault section or not, accurately searching an impassable section from the initial fault section, and avoiding the impassable section and an unavailable parking space when a preset route is planned. The system can timely and effectively monitor the impassable sections in the parking lot, timely and effective path planning is carried out on the vehicles in the parking lot, the response speed is high, the time is short, and the system is especially suitable for indoor parking lots with large traffic flow and complex roads.

The invention discloses a time-space domain obstacle detection method and system for a complex road scene, and the method comprises the steps: collecting point cloud data and a detection result of a historical frame, and carrying out the preprocessing of the point cloud data; detecting the preprocessed point cloud data by adopting a deep learning method to obtain known type target information, and detecting by adopting a grid occupying method to obtain unknown type target information; performing spatial domain association fusion on the known type target and the unknown type target to obtain a detection result of the current frame; performing motion estimation on the detection result of the historical frame to obtain a motion estimation result of the historical frame; and performing time domain fusion based on the detection result of the current frame and the motion estimation result of the historical frame to update the detection result of the current frame target. According to the method, various obstacles of known types and unknown types appearing on the road can be better detected, and the method has better performance on detection of shielded targets, small-size targets and the like.

The invention discloses a method for optimizing and solving an intelligent unmanned vehicle path planning problem based on a multi-constraint correction CW algorithm. According to the method, a simpleCW algorithm is modified, different constraint conditions such as weight limit, time windows and vehicle types are added to existing path optimization, and an intelligent unmanned vehicle delivery route is planned. According to the method, firstly, modeling is carried out on actual problems, then a corrected CW algorithm is used for solving the constraint problems of weight limitation and a timewindow in delivery route planning, and then different processing is carried out on planned routes according to different transport task optimization types, so that an optimal unmanned vehicle deliveryscheme is found.

The invention discloses a robot obstacle avoidance method and system based on multi-sensor fusion estimation and iterative pre-search, and the method comprises the steps: obtaining a pose measurement value of a robot according to the odometer data, position data and obstacle sensing data of the movement of the robot; according to the three-axis angular velocity and the three-axis acceleration of the motion of the robot, a pose prediction value of the robot is obtained; the pose of the robot is obtained according to the pose measurement value and the pose prediction value of the robot; judging whether the obstacle is located on a linear path of the current pose and the target point of the robot, if so, searching a motion path of the robot along the edge of the obstacle, respectively calculating cost values for the motion path, and taking the motion path with the lowest cost value as a middle target point, and carrying out calculation on the robot to the middle target point; and performing path planning from the middle target point to the target point. And real-time obstacle avoidance of the robot is realized.