WO2018176596A1 - 基于权重改进粒子群算法的无人自行车路径规划方法 - Google Patents

基于权重改进粒子群算法的无人自行车路径规划方法 Download PDF

Info

Publication number
WO2018176596A1
WO2018176596A1 PCT/CN2017/084510 CN2017084510W WO2018176596A1 WO 2018176596 A1 WO2018176596 A1 WO 2018176596A1 CN 2017084510 W CN2017084510 W CN 2017084510W WO 2018176596 A1 WO2018176596 A1 WO 2018176596A1
Authority
WO
WIPO (PCT)
Prior art keywords
path
unmanned
point
unmanned bicycle
path planning
Prior art date
Application number
PCT/CN2017/084510
Other languages
English (en)
French (fr)
Inventor
吴建国
Original Assignee
深圳市靖洲科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市靖洲科技有限公司 filed Critical 深圳市靖洲科技有限公司
Publication of WO2018176596A1 publication Critical patent/WO2018176596A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0214Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0217Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with energy consumption, time reduction or distance reduction criteria
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/004Artificial life, i.e. computing arrangements simulating life
    • G06N3/006Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]

Definitions

  • the invention relates to an unmanned bicycle technology, in particular to an unmanned bicycle path planning method based on a weight improved particle swarm optimization algorithm.
  • Baidu has announced the development of a complex artificial intelligence unmanned bicycle.
  • This product is an unmanned bicycle with complex artificial intelligence such as environmental awareness, planning and self-balancing control. It mainly integrates Baidu in artificial intelligence.
  • the achievements of deep learning, big data and cloud computing technologies however, there is no disclosure of technical details.
  • most of the sports intervention service systems with wide coverage, low cost and high specificity are adopted, and the intervention of the unmanned bicycles in accordance with the actual situation is expected to solve the problem of bicycle obstacle avoidance.
  • the obstacle avoidance path planning system determines how the vehicle reaches the target position under various constraints and path obstacle conditions, including environmental constraints embodied in safety, and systemic kinematic constraints embodying feasibility.
  • System dynamics constraints that reflect ride and stability, as well as specific optimization index constraints, such as the shortest time or shortest distance.
  • optimization index constraints such as the shortest time or shortest distance.
  • the global path planning problem is equivalent to the problem of path generation between the starting point and the end point. Solving the global path planning problem generally requires the completion of the typical road and its digital storage in advance. Way, that is, environmental map, when environmental changes or other factors lead to planning results When it is not feasible, you need to restart the global plan to get a new feasible path to continue exercising.
  • PSO Particle swarm optimization
  • the object of the present invention is to provide an unmanned bicycle path planning method based on the weight improved particle swarm optimization algorithm, which comprises the following steps:
  • the step (1) is embodied as follows: in the three-dimensional space coordinate system, the sequence coordinates of the path point are three-dimensional, and coordinate coordinates are performed for effective coding, and the starting point of the unmanned bicycle is in the new coordinate system.
  • S is the origin
  • the x-axis is the line connecting the origin and the target point G
  • the y-axis is perpendicular to the x-axis
  • the y-axis is parallel to the horizontal plane
  • the z-axis is perpendicular to the origin and perpendicular to the xoy plane
  • the coordinate transformation and inverse transformation formula are:
  • is the angle between the x'oy' projection on the x'oy' projection and the x'oy' plane x' axis
  • is the angle between the x-axis and its projection at x'oy', where xoy is the plane obtained after the transformation.
  • the minimum z-axis that the unmanned bicycle can reach is z min
  • the minimum value of y is y min
  • the maximum value is y max , whose coordinate point is o'(y min , y max ), moves the o point in the coordinate system yoz to o", and at the same point, do o "y" parallel to oy, do o"z" parallel to Oz
  • , z i ′′ z i +
  • the step (2) comprises establishing a fitness function of the static barrier, establishing a fitness function of the dynamic barrier, and establishing an adaptation of the integrated barrier.
  • the static barrier of the step (2) requires the unmanned bicycle to always be in a safe collision-free environment during the determined static environment, from the starting point to the end point, that is, the collision function in each of the broken line segments.
  • the radius of the bicycle expands and the unmanned bicycle is processed according to the mass point.
  • 0 and 1,0 are obstacles in the range of the line segment, and 1 is the obstacle-free object.
  • the fitness function of the static barrier is expressed as:
  • the dynamic barrier fitness function of the step (2) requires the unmanned bicycle to be in the process of having the other unmanned bicycles participate in the operation, and the unmanned bicycle is in the process of riding from the starting point to the end point, the unmanned bicycle and the unmanned bicycle
  • the distance of any other unmanned bicycle is greater than the sum of the safety radii of the two.
  • the position and speed of other unmanned bicycles can be obtained by the control terminal of the entire bicycle control system, each cycle
  • the path planning algorithm generates real-time optimal path solutions based on the position and velocity of other unmanned bicycles.
  • the trajectories of other unmanned bicycles are regarded as a uniform linear motion of the current time speed.
  • the dynamic barrier fitness function of a human bicycle is:
  • R 0 is the safe radius of the unmanned bicycle and R k is the safe radius of the obstacle.
  • the path length is equal to the length of the determined points on the plane, and the objective function is:
  • the comprehensive fitness function of the unmanned bicycle path planning algorithm based on the weighted improved particle swarm optimization algorithm can be obtained:
  • the bicycle can be driven strictly according to the planned path, and the optimal path can be generated in both the dynamic and static environments, the parameters to be adjusted are small, the model is simple, and the calculation is convenient.
  • FIG. 1 is a flow chart of an unmanned bicycle path planning method based on a weight improved particle swarm optimization algorithm according to an embodiment of the present invention
  • FIG. 2 is a diagram showing a simulation effect of a static obstacle according to an embodiment of the present invention
  • FIG. 3 is a diagram showing a simulation effect of a dynamic obstacle according to an embodiment of the present invention.
  • FIG. 1 includes the following steps:
  • step (1) is specifically implemented as: in the three-dimensional space coordinate system, the coordinate of the path point sequence is three-dimensional, and coordinate coordinates are performed for effective coding, and in the new coordinate system, the unmanned bicycle starting point S
  • the x-axis is the line connecting the origin and the target point G.
  • the y-axis is perpendicular to the x-axis
  • the y-axis is parallel to the horizontal plane
  • the z-axis is perpendicular to the origin and perpendicular to the xoy plane.
  • the coordinate transformation and inverse transformation formula are:
  • is the angle between the x'oy' projection on the x'oy' projection and the x'oy' plane x' axis
  • is the angle between the x-axis and its projection at x'oy', where xoy is the plane obtained after the transformation.
  • the minimum z-axis that the unmanned bicycle can reach is z min
  • the minimum value of y is y min
  • the maximum value is y max .
  • the coordinate point is o'(y min , y max ), move the o point in the coordinate system yoz to o", and at the same time do o "y" parallel to oy, do o"z" parallel to oz, then new
  • , z i ′′ z i +
  • Step (2) includes establishing a fitness function of the static barrier, establishing a fitness function of the dynamic barrier, and establishing an adaptation of the integrated barrier.
  • the static barrier requires the unmanned bicycle to be in a safe and collision-free environment from the starting point to the end point in the determined static environment, that is, the collision function of each broken line segment is multiplied by 1, in n
  • the plane is divided into n+1 spaces, and the final path is composed of n+1 folding line segments.
  • the passing area of each folding line segment is detected to determine whether there is an obstacle, and the obstacle is expanded according to the radius of the unmanned bicycle.
  • the range has obstacles, and 1 is no obstacle.
  • the static barrier fitness function is expressed as:
  • the dynamic barrier fitness function requires unmanned bicycles to have more than two distances between the unmanned bicycle and any other unmanned bicycle during the process of the unmanned bicycle from the starting point to the end of the operation.
  • the path planning algorithm is based on other The position and speed of the unmanned bicycle produce a real-time optimal path solution.
  • the trajectory of other unmanned bicycles is regarded as a uniform linear motion of the current time speed, and the dynamic barrier adaptive function of the unmanned bicycle for:
  • the final goal of the 3D path planning algorithm is that the path length is equal to the determined length of each point on the plane.
  • the objective function is:
  • the comprehensive fitness function of the unmanned bicycle path planning algorithm based on the weighted improved particle swarm optimization algorithm can be obtained:
  • the inverse transformation formula is: Step 5: Update the position of the particle, calculate the multiple weight values of the particle, and make a defense constraint, replacing the value beyond the range of (z y, , min , z y, max ) with the nearest one; Step: Update the optimal fitness and global optimal fitness of each particle; Step 7: Go to the fifth step and iterate until after reaching the maximum number of iterations or after achieving the required precision, smoothing by Schumann filtering Processing, modifying the calculated path, displaying the calculation result and the optimal path.
  • static obstacle simulation the particle population size is 100, the particle dimension is 10, the maximum number of iterations is 100, and the shortest path is 10, and the unmanned bicycle starts at One long In a narrow environment with a degree of 10 and a middle width of 2, the positions of the remaining obstacles can be randomly set.
  • the simulation diagram is shown in Figure 2.
  • the particle population size is 100
  • the particle dimension is 10
  • the maximum number of iterations is 100
  • the shortest path is 10.125.
  • the initial unmanned bicycle and dynamic obstacle parameters are shown in Table 1.
  • the simulation result table shows that the path planning method of the present invention enables the bicycle to travel strictly according to the planned path and can be in both dynamic and static environments.
  • the optimal path can be generated, the parameters to be adjusted are small, the model is simple, and the calculation is convenient.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Theoretical Computer Science (AREA)
  • Biophysics (AREA)
  • Mathematical Physics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

基于权重改进粒子群算法的无人自行车路径规划方法,包括如下步骤:(1)根据无人自行车的工作环境,进行环境建模和编码;(2)建立适应度函数;(3)基于权重改进粒子群算法进行无人自行车的路径规划。该路径规划方法,可使得自行车严格按照规划路径行驶,并且可以在动态和静态环境中都可以产生最优路径,需要调整的参数少,模型简单,计算方便。

Description

基于权重改进粒子群算法的无人自行车路径规划方法 技术领域
本发明涉及无人自行车技术,特别是一种基于权重改进粒子群算法的无人自行车路径规划方法。
背景技术
自20世纪60年代移动机器人诞生以来,研究人员一直梦想研究无人智能交通工具,作为智能交通系统的重要组成部分,无人自行车排除了人为不确定因素的影响,不仅可以提高驾驶安全性,而且可以解决交通拥堵,提高能源利用率,百度曾宣布开发复杂人工智能无人自行车,该产品是具备环境感知、规划和自平衡控制等复杂人工智能的无人自行车,主要集合了百度在人工智能、深度学习、大数据和云计算技术的成就,然而对技术细节没有任何披露。目前大多采用采用覆盖面广、成本低,且针对性强的运动干预服务系统,对无人自行车的运动进行符合实际情况的干预,有望解决自行车避障等问题。
作为无人自行车的智能核心,避障路径规划系统决定车辆如何在多种约束条件和路径障碍物条件下到达目标位置,这些约束包括体现为安全性的环境约束,体现可行性的系统运动学约束,体现平顺性和稳定性的系统动力学约束以及特定的优化指标约束,如最短时间或最短距离等。在无人自行车应用中,这些约束集中在全局路径规划中得到满足,全局路径规划问题等同于起点和终点间路径生成的问题,解决全局路径规划问题一般要求提前获知完成的典型道路及其数字化存储方式,也就是环境地图,当环境变化或其他因素导致规划结果 不可行时,需要重启全局规划得到新的可行路径才能继续行使。
粒子群优化算法是一种新的群体智能优化算法,这种算法易于实现、参数可控,然而还存在很多问题,比如收敛性差等问题,因此需要提出一种新的粒子群方法,加入改进的权重系数,通过对坐标进行变换,建立新的地图,同时针对动态和静态的障碍物搭配相应的适应度函数进行壁障,并通过仿真试验对算法的有效性进行验证。
发明内容
本发明的目的在于提供一种基于权重改进粒子群算法的无人自行车路径规划方法,包括如下步骤:
(1)根据无人自行车的工作环境,进行环境建模和编码;
(2)建立适应度函数;
(3)基于权重改进粒子群算法进行无人自行车的路径规划。
优选的,步骤(1)具体实施为:在三维空间坐标系中,路径点序列坐标为三维的,为了进行有效编码,对坐标进行坐标变换,在新的坐标系中,以无人自行车起始点S为原点,x轴为原点与目标点G的连线,y轴垂直于x轴,y轴与水平面平行,z轴过原点且垂直于xoy平面,其中坐标变换和反变换公式为:
Figure PCTCN2017084510-appb-000001
其中:α为x轴在x’oy’投影与x’oy’平面x’轴的夹角,β为x轴与其在x’oy’投影的夹角,其中xoy为变换后得到的平面。
优选的,所述步骤(1)中所述编码的方式为将原点与目标点的线段分成n等分,等分点为xi(i=1,2,...,n),分别过xi做平面垂直于x轴,xi点对应的平面为Qi,随机在Qi上选择一点pi,共产生了n个路径点,形成一条随机路径,将路径点 简化为二维的(yi,zi)坐标,然后设置决定平面取点精度的参数φ,在YOZ坐标系中无人自行车能到达的z轴最小值为zmin,y的最小值为ymin,最大值为ymax,其坐标点为o′(ymin,ymax),将坐标系yoz中的o点移动到o”,同时在该点做o”y”平行于oy,做o”z”平行于oz,则新的坐标为:yi″=yi+|ymin|,zi″=zi+|zmin|,然后对于每一个实数zyi=φ2×zi″×(ymax-ymin)+φ2×yi′,直接采用实数的编码方式zy={S,zy1,zy2,zy3,...,zyn...}。
优选的,所述步骤(2)包括建立静态壁障的适应度函数,建立动态壁障的适应度函数以及建立综合壁障的适应度。
优选的,所述步骤(2)的静态壁障要求无人自行车在确定的静态环境下,从起点骑行到终点的过程中始终在安全无碰撞环境内,也就是在各折线段的碰撞函数相乘为1,在n个平面内划分n+1个空间,最终路径由n+1条折线段构成,检测每条折线段的经过区域,判断是否有障碍物,同时对障碍物按照无人自行车半径进性膨胀,无人自行车按照质点来处理,设碰撞检测函数为S(i)=s(y′i-1,z′i-1,y′i,z′i),函数只返回0和1,0为折线段范围有障碍物,1为无障碍物,则静态壁障适应度函数表示为:
Figure PCTCN2017084510-appb-000002
优选的,所述步骤(2)的动态壁障适应度函数要求无人自行车在有其他无人自行车共同参与作业情况下,无人自行车从起点骑行到的终点的过程中,无人自行车与任意其他无人自行车的距离大于两者的安全半径之和,一个无人自行车在执行骑行任务的时候,其他无人自行车的位置和速度可以由整个自行车控制系统的控制终端得到,每个循环控制期内,路径规划算法根据其他无人自行车的位置和速度产生实时最优路径解,在整个路径规划过程中,将其他无人自行车的轨迹视为当前时刻速度的一个匀速直线运动,则无人自行车的动态壁障适应度函数为:
Figure PCTCN2017084510-appb-000003
其中R0为无人自行车的安全半径,Rk为障碍物的安全半径。
优选的,根据路径最短为三维路径规划算法最终目标,路径长度等于平面上已确定的各点连线的长度,目标函数为:
Figure PCTCN2017084510-appb-000004
将静态适应度函数、动态适应度函数以及距离适应度函数有机综合在一起,可以得到基于权重改进粒子群算法的无人自行车路径规划算法的综合适应度函数:
f=f1×f2×f3  (5)。
优选的,步骤(3)得算法流程为:第一步:建模设置参数by=(ymax-ymin),bz=(zmax-zmin),zy,max=by×bz,zmin=0,其中by,bz分别是新坐标y轴和z轴的上界,粒子搜索范围被限制在(zy,,min,zy,max);第二步:把所有静止状态的障碍物按无人自行车半径进行膨胀,确保行进安全;第三步:将所有粒子初始化于无障碍物路径,以使在以后的搜索过程中更快找到最优路径;第四步:将产生的粒子在平面内进行反变换,并代入适应度函数进行适应度的计算,计算出粒子初始化时的适应值,并保存到当前最短路径值中;第五步:更新粒子的位置,计算粒子的多个权重值,并进行辩解约束,将超出(zy,,min,zy,max)范围的值以最接近的一个边界值替代;第六步:更新每一个粒子的最优适应度和全局最优适应度;第七步:转到第五步进行迭代,直到到达最大迭代次数后或者达到需要的精度后,采用舒曼滤波法进行平滑处理,对算出的路径进行修改,显示计算结果与最优路径。
采用本发明的路径规划方法,可使得自行车严格按照规划路径行驶,并且可以在动态和静态环境中都可以产生最优路径,需要调整的参数少,模型简单,计算方便。
根据下文结合附图对本发明具体实施例的详细描述,本领域技术人员将会更加明了本发明的上述以及其他目的、优点和特征。
附图说明
后文将参照附图以示例性而非限制性的方式详细描述本发明的一些具体实施例。附图中相同的附图标记标示了相同或类似的部件或部分。本领域技术人员应该理解,这些附图未必是按比例绘制的。本发明的目标及特征考虑到如下结合附图的描述将更加明显,附图中:
图1为根据本发明实施例的基于权重改进粒子群算法的无人自行车路径规划方法流程图;
图2为根据本发明实施例的静态障碍物仿真效果图;
图3为根据本发明实施例的动态障碍物仿真效果图。
具体实施方式
结合附图如下详细说明一种的基于权重改进粒子群算法的无人自行车路径规划方法,如图1所示包括如下步骤:
(1)根据无人自行车的工作环境,进行环境建模和编码;
(2)建立适应度函数;
(3)基于权重改进粒子群算法进行无人自行车的路径规划。
其中,步骤(1)具体实施为:在三维空间坐标系中,路径点序列坐标为三维的,为了进行有效编码,对坐标进行坐标变换,在新的坐标系中,以无人自行车起始点S为原点,x轴为原点与目标点G的连线,y轴垂直于x轴,y轴与水平面平行,z轴过原点且垂直于xoy平面,其中坐标变换和反变换公式为:
Figure PCTCN2017084510-appb-000005
其中:α为x轴在x’oy’投影与x’oy’平面x’轴的夹角,β为x轴与其在x’oy’投影的夹角,其中xoy为变换后得到的平面。
其中,步骤(1)中的编码的方式为将原点与目标点的线段分成n等分,等分点为xi(i=1,2,...,n),分别过xi做平面垂直于x轴,xi点对应的平面为Qi,随机在Qi上选择一点pi,共产生了n个路径点,形成一条随机路径,将路径点简化为二维的(yi,zi)坐标,然后设置决定平面取点精度的参数φ,在YOZ坐标系中无人自行车能到达的z轴最小值为zmin,y的最小值为ymin,最大值为ymax,其坐标点为o′(ymin,ymax),将坐标系yoz中的o点移动到o”,同时在该点做o”y”平行于oy,做o”z”平行于oz,则新的坐标为:yi″=yi+|ymin|,zi″=zi+|zmin|,然后对于每一个实数zyi=φ2×zi″×(ymax-ymin)+φ2×yi′,直接采用实数的编码方式zy={S,zy1,zy2,zy3,...,zyn...}。
而步骤(2)包括建立静态壁障的适应度函数,建立动态壁障的适应度函数以及建立综合壁障的适应度。其中,静态壁障要求无人自行车在确定的静态环境下,从起点骑行到终点的过程中始终在安全无碰撞环境内,也就是在各折线段的碰撞函数相乘为1,在n个平面内划分n+1个空间,最终路径由n+1条折线段构成,检测每条折线段的经过区域,判断是否有障碍物,同时对障碍物按照无人自行车半径进性膨胀,无人自行车按照质点来处理,设碰撞检测函数为S(i)=s(y′i-1,z′i-1,y′i,z′i),函数只返回0和1,0为折线段范围有障碍物,1为无障碍物,则静态壁障适应度函数表示为:
Figure PCTCN2017084510-appb-000006
而动态壁障适应度函数要求无人自行车在有其他无人自行车共同参与作业情况下,无人自行车从起点骑行到的终点的过程中,无人自行车与任意其他无人自行车的距离大于两者的安全半径之和,一个无人自行车在执行骑行任务的时候,其他无人自行车的位置和速度可以由整个自行车控制系统的控制终端得到,每个循环控制期内,路径规划算法根据其他无人自行车的位置和速度产生实时最优路径解,在整个路径规划过程中,将其他无人自行车的轨迹视为当前时刻速度的一个匀速直线运动,则无人自行车 的动态壁障适应度函数为:
Figure PCTCN2017084510-appb-000007
其中R0为无人自行车的安全半径,Rk为障碍物的安全半径。根据路径最短为三维路径规划算法最终目标,路径长度等于平面上已确定的各点连线的长度,目标函数为:
Figure PCTCN2017084510-appb-000008
将静态适应度函数、动态适应度函数以及距离适应度函数有机综合在一起,可以得到基于权重改进粒子群算法的无人自行车路径规划算法的综合适应度函数:
f=f1×f2×f3  (5)。
对于步骤(3)的算法流程,分解为:第一步:建模设置参数by=(ymax-ymin),bz=(zmax-zmin),zy,max=by×bz,zmin=0,其中by,bz分别是新坐标y轴和z轴的上界,粒子搜索范围被限制在(zy,,min,zy,max);第二步:把所有静止状态的障碍物按无人自行车半径进行膨胀,确保行进安全;第三步:将所有粒子初始化于无障碍物路径,以使在以后的搜索过程中更快找到最优路径;第四步:将产生的粒子在平面内进行反变换,并代入适应度函数进行适应度的计算,计算出粒子初始化时的适应值,并保存到当前最短路径值中,反变换公式为:
Figure PCTCN2017084510-appb-000009
第五步:更新粒子的位置,计算粒子的多个权重值,并进行辩解约束,将超出(zy,,min,zy,max)范围的值以最接近的一个边界值替代;第六步:更新每一个粒子的最优适应度和全局最优适应度;第七步:转到第五步进行迭代,直到到达最大迭代次数后或者达到需要的精度后,采用舒曼滤波法进行平滑处理,对算出的路径进行修改,显示计算结果与最优路径。
对于该算法在MATLAB平台进行仿真,选择原始地图大小为20*10*10,S(0,0,0)和G(20,5,5)分别为起始点和终点,参数φ=100,仿真分为静态障碍物和动态障碍物2个部分,对于静态障碍物仿真,粒子种群大小为100,粒子维数为10,最大迭代次数为100,最后得到最短路径为10,无人自行车起始在一个长 度为10,中部宽度为2的狭窄环境中,其余障碍物的位置可以随机设置。仿真图如图2所示,而在动态障碍物环境下,粒子种群大小为100,粒子维数为10,最大迭代次数为100,最后得到最短路径为10.125。初始时无人自行车和动态障碍物参数如表1所示,
物体类别 当前位置 速度(cm/s) 安全半径 运动轨迹
无人自行车1 (0,0,0) 10 1 待求
目的地 (20,5,5) 0 0
障碍物1 (3,5,0) 5 1 垂直于z轴正
障碍物2 (5,10,5) 5 1 平行于x轴正
障碍物3 (18,0,0) 12 1 对角线方向
那么从初始到最终状态的4个不同状态的最后路径如图3所示,仿真结果表表明:本发明的路径规划方法,可使得自行车严格按照规划路径行驶,并且可以在动态和静态环境中都可以产生最优路径,需要调整的参数少,模型简单,计算方便。
虽然本发明已经参考特定的说明性实施例进行了描述,但是不会受到这些实施例的限定而仅仅受到附加权利要求的限定。本领域技术人员应当理解可以在不偏离本发明的保护范围和精神的情况下对本发明的实施例能够进行改动和修改。

Claims (8)

  1. 基于权重改进粒子群算法的无人自行车路径规划方法,其特征在于:包括如下步骤:
    (1)根据无人自行车的工作环境,进行环境建模和编码;
    (2)建立适应度函数;
    (3)基于权重改进粒子群算法进行无人自行车的路径规划。
  2. 根据权利要求1的路径规划方法,其特征在于:所述步骤(1)具体实施为:在三维空间坐标系中,路径点序列坐标为三维的,为了进行有效编码,对坐标进行坐标变换,在新的坐标系中,以无人自行车起始点S为原点,x轴为原点与目标点G的连线,y轴垂直于x轴,y轴与水平面平行,z轴过原点且垂直于xoy平面,其中坐标变换和反变换公式为:
    Figure PCTCN2017084510-appb-100001
    其中:α为x轴在x’oy’投影与x’oy’平面x’轴的夹角,β为x轴与其在x’oy’投影的夹角,其中xoy为变换后得到的平面。
  3. 根据权利要求2的路径规划方法,其特征在于:步骤(1)中所述编码的方式为将原点与目标点的线段分成n等分,等分点为xi(i=1,2,...,n),分别过xi做平面垂直于x轴,xi点对应的平面为Qi,随机在Qi上选择一点pi,共产生了n个路径点,形成一条随机路径,将路径点简化为二维的(yi,zi)坐标,然后设置决定平面取点精度的参数φ,在YOZ坐标系中无人自行车能到达的z轴最小值为zmin,y的最小值为ymin,最大值为ymax,其坐标点为o′(ymin,ymax),将坐标系yoz中的o点移动到o”,同时在该点做o”y”平行于oy,做o”z”平行于oz,则新的坐标为:yi″=yi+|ymin|,zi″=zi+|zmin|,然后对于每一个实数zyi=φ2×zi″×(ymax-ymin)+φ2×yi′,直接采用实数的编码方式zy={S,zy1,zy2,zy3,...,zyn...}。
  4. 根据权利要求1的路径规划方法,其特征在于:步骤(2)包括建立静态壁障的适应度函数,建立动态壁障的适应度函数以及建立综合壁障的适应度。
  5. 根据权利要求4的路径规划方法,其特征在于:步骤(2)的静态壁障要求无人自行车在确定的静态环境下,从起点骑行到终点的过程中始终在安全无碰撞环境内,也就是在各折线段的碰撞函数相乘为1,在n个平面内划分n+1个空间,最终路径由n+1条折线段构成,检测每条折线段的经过区域,判断是否有障碍物,同时对障碍物按照无人自行车半径进性膨胀,无人自行车按照质点来处理,设碰撞检测函数为S(i)=s(y′i-1,z′i-1,y′i,z′i),函数只返回0和1,0为折线段范围有障碍物,1为无障碍物,则静态壁障适应度函数表示为:
    Figure PCTCN2017084510-appb-100002
  6. 根据权利要求4的路径规划方法,其特征在于:步骤(2)的动态壁障适应度函数要求无人自行车在有其他无人自行车共同参与作业情况下,无人自行车从起点骑行到的终点的过程中,无人自行车与任意其他无人自行车的距离大于两者的安全半径之和,一个无人自行车在执行骑行任务的时候,其他无人自行车的位置和速度可以由整个自行车控制系统的控制终端得到,每个循环控制期内,路径规划算法根据其他无人自行车的位置和速度产生实时最优路径解,在整个路径规划过程中,将其他无人自行车的轨迹视为当前时刻速度的一个匀速直线运动,则无人自行车的动态壁障适应度函数为:
    Figure PCTCN2017084510-appb-100003
    其中R0为无人自行车的安全半径,Rk为障碍物的安全半径。
  7. 根据权利要求4的路径规划方法,其特征在于:根据路径最短为三维路径规划算法最终目标,路径长度等于平面上已确定的各点连线的长度,目标函数为:
    Figure PCTCN2017084510-appb-100004
    将静态适应度函数、动态适应度函数以及距离适应度函数有机综合在一起,可以得到所述基于权重改进粒 子群算法的无人自行车路径规划算法的综合适应度函数:
    f=f1×f2×f3 (5)。
  8. 根据权利要求1的路径规划方法,其特征在于:所述步骤(3)的算法流程为:第一步:建模设置参数by=(ymax-ymin),bz=(zmax-zmin),zy,max=by×bz,zmin=0,其中by,bz分别是新坐标y轴和z轴的上界,粒子搜索范围被限制在(zy,,min,zy,max);第二步:把所有静止状态的障碍物按无人自行车半径进行膨胀,确保行进安全;第三步:将所有粒子初始化于无障碍物路径,以使在以后的搜索过程中更快找到最优路径;第四步:将产生的粒子在平面内进行反变换,并代入适应度函数进行适应度的计算,计算出粒子初始化时的适应值,并保存到当前最短路径值中;第五步:更新粒子的位置,计算粒子的多个权重值,并进行辩解约束,将超出(zy,,min,zy,max)范围的值以最接近的一个边界值替代;第六步:更新每一个粒子的最优适应度和全局最优适应度;第七步:转到第五步进行迭代,直到到达最大迭代次数后或者达到需要的精度后,采用舒曼滤波法进行平滑处理,对算出的路径进行修改,显示计算结果与最优路径。
PCT/CN2017/084510 2017-03-31 2017-05-16 基于权重改进粒子群算法的无人自行车路径规划方法 WO2018176596A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710210985.4 2017-03-31
CN201710210985.4A CN106843236A (zh) 2017-03-31 2017-03-31 基于权重改进粒子群算法的无人自行车路径规划方法

Publications (1)

Publication Number Publication Date
WO2018176596A1 true WO2018176596A1 (zh) 2018-10-04

Family

ID=59141228

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/084510 WO2018176596A1 (zh) 2017-03-31 2017-05-16 基于权重改进粒子群算法的无人自行车路径规划方法

Country Status (2)

Country Link
CN (1) CN106843236A (zh)
WO (1) WO2018176596A1 (zh)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109799822A (zh) * 2019-01-30 2019-05-24 中国石油大学(华东) 移动机器人全局平滑路径规划方法
CN111198576A (zh) * 2020-01-09 2020-05-26 哈尔滨工程大学 人工智能大数据下粒子化无人机组控制方法、介质及机组
CN113205173A (zh) * 2021-05-28 2021-08-03 中国工商银行股份有限公司 一种电子券选择方法和装置
CN113255871A (zh) * 2021-05-11 2021-08-13 合肥宏晶微电子科技股份有限公司 路径规划方法、电子设备、计算机可读存储介质
CN114740873A (zh) * 2022-04-13 2022-07-12 青岛科技大学 一种基于多目标改进粒子群算法的自主式水下机器人的路径规划方法
CN115016508A (zh) * 2022-07-29 2022-09-06 陕西师范大学 基于区域分割的多目标粒子群优化算法的机器人路径规划方法
CN115237117A (zh) * 2022-04-28 2022-10-25 安徽理工大学 基于变异自适应粒子群算法的移动机器人路径规划方法
CN117592633A (zh) * 2024-01-18 2024-02-23 深圳金智凌轩视讯技术有限公司 一种智慧巡检优化方法
CN117891258A (zh) * 2024-03-12 2024-04-16 江苏韦尔汀轨道工程技术有限公司 一种轨道焊接路径智能规划方法
CN115016508B (zh) * 2022-07-29 2024-07-05 陕西师范大学 基于区域分割的多目标粒子群优化算法的机器人路径规划方法

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10571921B2 (en) * 2017-09-18 2020-02-25 Baidu Usa Llc Path optimization based on constrained smoothing spline for autonomous driving vehicles
CN108153310B (zh) * 2017-12-22 2020-11-13 南开大学 一种基于人类行为模拟的移动机器人实时运动规划方法
CN107992051B (zh) * 2017-12-26 2020-10-23 江南大学 基于改进的多目标粒子群算法的无人车路径规划方法
CN109000653B (zh) * 2018-06-06 2020-05-05 西南交通大学 一种多维空间多载运工具路径智能优化方法
CN111459152A (zh) * 2019-01-21 2020-07-28 珠海格力电器股份有限公司 智能设备的控制方法、控制装置和空气净化器
CN110782092A (zh) * 2019-10-25 2020-02-11 北京京东乾石科技有限公司 无人配送车在非结构化场景中的轨迹规划方法及装置
CN110928297B (zh) * 2019-10-28 2020-10-30 中南大学 基于多目标动态粒子群优化的智能公交车辆路径规划方法
CN111061276B (zh) * 2019-12-31 2022-07-26 芜湖哈特机器人产业技术研究院有限公司 基于动态区域划分的移动机器人重定位方法
CN111736618B (zh) * 2020-06-28 2021-08-10 清华大学 一种无人摩托转向控制参数整定方法和装置
CN113885501B (zh) * 2021-10-09 2024-05-03 南京晨光集团有限责任公司 一种用于无人车位姿调节的轨迹规划方法
CN115086229B (zh) * 2022-04-29 2023-07-11 珠海高凌信息科技股份有限公司 一种基于进化算法的sdn网络多路径计算方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101604166A (zh) * 2009-07-10 2009-12-16 杭州电子科技大学 一种基于粒子群优化算法的移动机器人路径规划方法
US20100121517A1 (en) * 2008-11-10 2010-05-13 Electronics And Telecommunications Research Institute Method and apparatus for generating safe path of mobile robot
CN101738195A (zh) * 2009-12-24 2010-06-16 厦门大学 基于环境建模与自适应窗口的移动机器人路径规划方法
CN102129249A (zh) * 2011-01-10 2011-07-20 中国矿业大学 一种危险源环境下的机器人全局路径规划方法
CN102854880A (zh) * 2012-10-08 2013-01-02 中国矿业大学 面向混合地形区域不确定环境的机器人全局路径规划方法
CN103557867A (zh) * 2013-10-09 2014-02-05 哈尔滨工程大学 一种基于稀疏a*搜索的三维多uav协同航迹规划方法
CN103901887A (zh) * 2014-03-04 2014-07-02 重庆邮电大学 一种基于改进粒子群算法的多移动机器人编队控制方法

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101436073A (zh) * 2008-12-03 2009-05-20 江南大学 基于量子行为粒子群算法的轮式移动机器人轨迹跟踪方法
US9373960B2 (en) * 2013-03-13 2016-06-21 Oracle International Corporation Computerized system and method for distributed energy resource scheduling
CN103646278B (zh) * 2013-11-14 2018-04-24 扬州西岐自动化科技有限公司 基于自适应策略的粒子群算法在机器人路径规划中的应用
WO2015168431A1 (en) * 2014-04-30 2015-11-05 Stc.Unm Optimization methods for radiation therapy planning
CN105387875B (zh) * 2015-12-24 2018-01-12 安徽工程大学 基于蚁群算法的移动机器人路径规划方法的一种改进

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100121517A1 (en) * 2008-11-10 2010-05-13 Electronics And Telecommunications Research Institute Method and apparatus for generating safe path of mobile robot
CN101604166A (zh) * 2009-07-10 2009-12-16 杭州电子科技大学 一种基于粒子群优化算法的移动机器人路径规划方法
CN101738195A (zh) * 2009-12-24 2010-06-16 厦门大学 基于环境建模与自适应窗口的移动机器人路径规划方法
CN102129249A (zh) * 2011-01-10 2011-07-20 中国矿业大学 一种危险源环境下的机器人全局路径规划方法
CN102854880A (zh) * 2012-10-08 2013-01-02 中国矿业大学 面向混合地形区域不确定环境的机器人全局路径规划方法
CN103557867A (zh) * 2013-10-09 2014-02-05 哈尔滨工程大学 一种基于稀疏a*搜索的三维多uav协同航迹规划方法
CN103901887A (zh) * 2014-03-04 2014-07-02 重庆邮电大学 一种基于改进粒子群算法的多移动机器人编队控制方法

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109799822A (zh) * 2019-01-30 2019-05-24 中国石油大学(华东) 移动机器人全局平滑路径规划方法
CN111198576A (zh) * 2020-01-09 2020-05-26 哈尔滨工程大学 人工智能大数据下粒子化无人机组控制方法、介质及机组
CN113255871A (zh) * 2021-05-11 2021-08-13 合肥宏晶微电子科技股份有限公司 路径规划方法、电子设备、计算机可读存储介质
CN113205173A (zh) * 2021-05-28 2021-08-03 中国工商银行股份有限公司 一种电子券选择方法和装置
CN114740873A (zh) * 2022-04-13 2022-07-12 青岛科技大学 一种基于多目标改进粒子群算法的自主式水下机器人的路径规划方法
CN114740873B (zh) * 2022-04-13 2024-05-28 青岛科技大学 一种基于多目标改进粒子群算法的自主式水下机器人的路径规划方法
CN115237117A (zh) * 2022-04-28 2022-10-25 安徽理工大学 基于变异自适应粒子群算法的移动机器人路径规划方法
CN115237117B (zh) * 2022-04-28 2024-05-07 安徽理工大学 基于变异自适应粒子群算法的移动机器人路径规划方法
CN115016508A (zh) * 2022-07-29 2022-09-06 陕西师范大学 基于区域分割的多目标粒子群优化算法的机器人路径规划方法
CN115016508B (zh) * 2022-07-29 2024-07-05 陕西师范大学 基于区域分割的多目标粒子群优化算法的机器人路径规划方法
CN117592633A (zh) * 2024-01-18 2024-02-23 深圳金智凌轩视讯技术有限公司 一种智慧巡检优化方法
CN117592633B (zh) * 2024-01-18 2024-03-26 深圳金智凌轩视讯技术有限公司 一种智慧巡检优化方法
CN117891258A (zh) * 2024-03-12 2024-04-16 江苏韦尔汀轨道工程技术有限公司 一种轨道焊接路径智能规划方法
CN117891258B (zh) * 2024-03-12 2024-05-28 江苏韦尔汀轨道工程技术有限公司 一种轨道焊接路径智能规划方法

Also Published As

Publication number Publication date
CN106843236A (zh) 2017-06-13

Similar Documents

Publication Publication Date Title
WO2018176596A1 (zh) 基于权重改进粒子群算法的无人自行车路径规划方法
Yang et al. A new robot navigation algorithm based on a double-layer ant algorithm and trajectory optimization
CN112859859B (zh) 一种基于三维障碍物体素对象映射的动态栅格地图更新方法
WO2018176593A1 (zh) 一种面向无人自行车的局部避障路径规划方法
WO2016045615A1 (zh) 机器人静态路径规划方法
Sudhakara et al. Trajectory planning of a mobile robot using enhanced A-star algorithm
CN111381600B (zh) 一种基于粒子群算法的uuv路径规划方法
CN106371445A (zh) 一种基于拓扑地图的无人车规划控制方法
CN110703768A (zh) 一种改进型动态rrt*的移动机器人运动规划方法
CN112539750B (zh) 一种智能运输车路径规划方法
CN109978272A (zh) 一种基于多个全向移动机器人的路径规划系统及方法
CN106705975A (zh) 一种基于万有引力搜索的水面无人艇航路规划方法
CN113985922A (zh) 一种多目标约束下的无人机分级式路径规划方法
Chen et al. Dynamic obstacle avoidance for UAVs using a fast trajectory planning approach
CN113391633A (zh) 一种面向城市环境的移动机器人融合路径规划方法
Wu et al. Real-time three-dimensional smooth path planning for unmanned aerial vehicles in completely unknown cluttered environments
Wang et al. Research on AGV task path planning based on improved A* algorithm
BAYGIN et al. PSO based path planning approach for multi service robots in dynamic environments
CN116661502B (zh) 一种智能农业无人机路径规划方法
Omar et al. 3D path planning for unmanned aerial vehicles using visibility line based method
CN116560408A (zh) 一种用于无人机时间最优轨迹的序列凸优化方法及系统
CN111596668B (zh) 基于逆向强化学习的移动机器人拟人化路径规划方法
Wang et al. AGRNav: Efficient and Energy-Saving Autonomous Navigation for Air-Ground Robots in Occlusion-Prone Environments
Guo et al. Online path planning for UAV navigation based on quantum particle swarm optimization
Qin et al. A path planning algorithm based on deep reinforcement learning for mobile robots in unknown environment

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17904346

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 20/01/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 17904346

Country of ref document: EP

Kind code of ref document: A1