CN107168324A - A kind of robot path planning method based on ANFIS fuzzy neural networks - Google Patents
A kind of robot path planning method based on ANFIS fuzzy neural networks Download PDFInfo
- Publication number
- CN107168324A CN107168324A CN201710429378.7A CN201710429378A CN107168324A CN 107168324 A CN107168324 A CN 107168324A CN 201710429378 A CN201710429378 A CN 201710429378A CN 107168324 A CN107168324 A CN 107168324A
- Authority
- CN
- China
- Prior art keywords
- robot
- neural network
- mrow
- fuzzy
- fuzzy neural
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 59
- 238000000034 method Methods 0.000 title claims abstract description 40
- 230000003044 adaptive effect Effects 0.000 claims abstract description 8
- 230000006870 function Effects 0.000 claims abstract description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 16
- 238000013459 approach Methods 0.000 claims description 12
- 230000006399 behavior Effects 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 4
- 230000000903 blocking effect Effects 0.000 claims description 3
- 238000005381 potential energy Methods 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 206010024769 Local reaction Diseases 0.000 abstract 1
- 238000010276 construction Methods 0.000 abstract 1
- 230000007613 environmental effect Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 210000000653 nervous system Anatomy 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000002922 simulated annealing Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0217—Control 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
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
- Feedback Control In General (AREA)
Abstract
本发明公开了一种基于ANFIS模糊神经网络的机器人路径规划方法，主要解决传统反应式导航中复杂陷阱路径往复和路径冗余的问题。其规划步骤是首先对移动机器人建立运动学模型；借助神经网络的自主学习功能和模糊理论的模糊推理能力，提出一种模糊神经网络的移动机器人导航控制器；基于自适应模糊神经网络结构，构建Takagi‑Sugeno型模糊推理系统并作为机器人局部反应控制的参考模型；该模糊神经网络控制器实时输出偏移角度和运行速度，在线调整移动机器人的偏移方向，使移动机器人能够无碰撞自动调节速度趋向目标；采用改进虚拟目标方法，选择机器人能够逃离陷阱状态的最优路径。
The invention discloses a robot path planning method based on the ANFIS fuzzy neural network, which mainly solves the problems of complex trap path reciprocation and path redundancy in traditional reactive navigation. The planning steps are to first establish a kinematics model for the mobile robot; with the help of the autonomous learning function of the neural network and the fuzzy reasoning ability of the fuzzy theory, a mobile robot navigation controller based on the fuzzy neural network is proposed; based on the adaptive fuzzy neural network structure, the construction The Takagi‑Sugeno type fuzzy reasoning system is used as a reference model for the local reaction control of the robot; the fuzzy neural network controller outputs the offset angle and running speed in real time, and adjusts the offset direction of the mobile robot online, so that the mobile robot can automatically adjust the speed without collision Approaching the goal; adopting the improved virtual goal method to select the optimal path for the robot to escape from the trap state.
Description
技术领域technical field
本发明属于机器人技术领域，特别是一种涉及移动机器人的路径规划，可用于各类移动机器人的自主导航。The invention belongs to the technical field of robots, and in particular relates to path planning of mobile robots, which can be used for autonomous navigation of various mobile robots.
背景技术Background technique
路径规划问题是移动机器人导航的关键技术之一，主要任务是在有障碍物的环境中，按照一定的性能指标，寻找一条从起点到目标点之间一条最优或者接近最优的无碰撞路径。根据机器人对环境信息感知程度的不同,路径规划分为两种:环境信息完全知道的全局路径规划和环境信息完全未知或局部未知的局部路径规划。全局路径规划一般离线进行,常用的方法主要有可视图法、栅格法、结构空间法、拓扑法、模拟退火法、遗传算法和蚁群算法等智能算法。局部路径规划常用的方法有人工势场法、模糊逻辑算法和神经网络法等。神经网络因容错性强与具有自适应学习的特点，可以更好地在非结构化环境下进行感知信息的分析与融合，而模糊控制具有逻辑推理能力，对处理结构化知识更为有效.然而反应式导航缺乏对环境的全局认识，易使机器人陷入局部陷阱而无法到达终点。针对这一问题，目前提出的有效方法有行为融合、虚目标、沿轮廓跟踪等方法，但行为融合方法需要计算各行为的权值，增加了系统复杂度；沿轮廓跟踪方法受障碍物形状、大小影响较大；虚目标在复杂环境下不易去除虚拟子目标及易产生冗余路径。The path planning problem is one of the key technologies of mobile robot navigation. The main task is to find an optimal or near-optimal collision-free path from the starting point to the target point in an environment with obstacles and according to certain performance indicators. . According to the difference of robot's awareness of environmental information, path planning can be divided into two types: global path planning with fully known environmental information and local path planning with completely unknown or partially unknown environmental information. Global path planning is generally carried out offline, and the commonly used methods mainly include intelligent algorithms such as visual map method, grid method, structure space method, topology method, simulated annealing method, genetic algorithm and ant colony algorithm. Commonly used methods for local path planning include artificial potential field method, fuzzy logic algorithm and neural network method. Due to its strong fault tolerance and self-adaptive learning characteristics, neural network can better analyze and integrate perceptual information in an unstructured environment, while fuzzy control has the ability of logical reasoning, which is more effective for processing structured knowledge. Reactive navigation lacks a global understanding of the environment, and it is easy for the robot to fall into a local trap and fail to reach the destination. To solve this problem, currently proposed effective methods include behavior fusion, virtual target, tracking along the contour, etc., but the behavior fusion method needs to calculate the weight of each behavior, which increases the complexity of the system; the tracking along the contour is limited by the shape of obstacles, The size has a great influence; virtual objects are not easy to remove virtual sub-objects and easily generate redundant paths in complex environments.
发明内容Contents of the invention
发明目的：为了解决现有技术中反应式导航中复杂陷阱路径往复和路径冗余的问题，本发明提供一种基于ANFIS模糊神经网络的机器人路径规划方法，该方法不仅能够减少逻辑推理工作量，而且能够摆脱机器人趋向目标运行中的陷阱状态。Purpose of the invention: In order to solve the problem of complex trap path reciprocation and path redundancy in reactive navigation in the prior art, the present invention provides a robot path planning method based on ANFIS fuzzy neural network, which can not only reduce the workload of logical reasoning, And it can get rid of the trap state in the running of the robot towards the target.
技术方案：为实现上述目的，本发明采用的技术方案为：Technical scheme: in order to achieve the above object, the technical scheme adopted in the present invention is:
一种基于ANFIS模糊神经网络的机器人路径规划方法，首先对移动机器人建立运动学模型；借助神经网络的自主学习功能和模糊理论的模糊推理能力，提出一种模糊神经网络的移动机器人导航的模糊神经网络控制器；其基于自适应模糊神经网络结构，构建Takagi-Sugeno型模糊推理系统并作为机器人局部反应控制的参考模型；将障碍物的距离和位置的相关信息作为模糊神经网络控制器的两个输入，模糊神经网络控制器实时输出机器人偏移角度和运行速度，通过模糊神经网络控制器在线调整移动机器人的偏移方向，使移动机器人能够无碰撞自动调节速度趋向目标。A robot path planning method based on the ANFIS fuzzy neural network. Firstly, a kinematics model is established for the mobile robot; with the help of the autonomous learning function of the neural network and the fuzzy reasoning ability of the fuzzy theory, a fuzzy neural network for mobile robot navigation is proposed. Network controller; based on the adaptive fuzzy neural network structure, a Takagi-Sugeno type fuzzy inference system is constructed and used as a reference model for robot local response control; the distance and position related information of obstacles are used as two of the fuzzy neural network controller Input, the fuzzy neural network controller outputs the robot's offset angle and running speed in real time, and adjusts the offset direction of the mobile robot online through the fuzzy neural network controller, so that the mobile robot can automatically adjust the speed to the target without collision.
优选的：通过模糊神经网络控制器输出值表示机器人移动角度和速度，越靠近障碍物时输出角度绝对值越大，速度绝对值越小；当前方没有障碍物时，机器人沿预设定方向前进；当前方有一个障碍物时，机器人逐渐接近障碍物，在一定范围内实时改变偏移角度和速度，使机器人缓慢绕离障碍物驶向目标；当前方有两个及其以上障碍物时，移动机器人在行进过程中对虚拟目标进行实时调整，即机器人沿所识别的最后一个障碍物前进并避开除此之外所有障碍物，选择一条远离障碍的最优路径趋向目标。Preferably: the robot movement angle and speed are represented by the fuzzy neural network controller output value, the closer to the obstacle, the larger the absolute value of the output angle, and the smaller the absolute value of the speed; when there is no obstacle in front, the robot advances in a preset direction ;When there is an obstacle in front, the robot gradually approaches the obstacle, and changes the offset angle and speed in real time within a certain range, so that the robot slowly moves away from the obstacle and drives towards the target; when there are two or more obstacles in front, The mobile robot adjusts the virtual target in real time during the process of moving, that is, the robot advances along the last obstacle identified and avoids all other obstacles, and chooses an optimal path away from the obstacle to approach the target.
优选的：模糊神经网络控制器利用LMS算法和最小二乘法来完成输入/输出数据对的建模，使得Takagi-Sugeno型模糊推理系统能模拟出希望或实际的输入/输出关系。Preferably: the fuzzy neural network controller uses the LMS algorithm and the least square method to complete the modeling of the input/output data pair, so that the Takagi-Sugeno type fuzzy reasoning system can simulate the desired or actual input/output relationship.
优选的：模糊神经网络控制器在学习时，根据系统实际输出值与期望输出值计算出学习误差，再通过LMS算法对系统的偏移角度和速度进行调整。Preferably: when the fuzzy neural network controller is learning, it calculates the learning error according to the actual output value and the expected output value of the system, and then adjusts the offset angle and speed of the system through the LMS algorithm.
所述对移动机器人建立运动学模型的方法如下：The method for establishing a kinematic model of a mobile robot is as follows:
步骤101，移动机器人通过本体携带的测距传感器测量障碍物的距离，其中，机器人当前坐标为(xr,yr)，目标点坐标为(xt,yt)，E是机器人当前位置(xr,yr)到目标点(xt,yt)的矢量，其模长和向量角表示为：Step 101, the mobile robot measures the distance of obstacles through the ranging sensor carried by the body, where the current coordinates of the robot are (x r , y r ), the coordinates of the target point are (x t , y t ), and E is the current position of the robot ( x r , y r ) to the target point (x t , y t ), its modulus length and vector angle are expressed as:
En为机器人在目标距离势场中的势能，当前机器人与目标点的夹角，根据机器人当前位置不断修正，始终指向目标位置，下标n表示具体时刻；E n is the potential energy of the robot in the target distance potential field, The angle between the current robot and the target point is constantly corrected according to the current position of the robot, and always points to the target position, and the subscript n indicates the specific moment;
步骤102，速度模型，移动机器人在导航任务中的速度由机器人与周围障碍物之间距离决定，当无障碍物阻挡时，机器人全速前进，当遇到障碍物时减速行驶，遵循以下公式：Step 102, speed model, the speed of the mobile robot in the navigation task is determined by the distance between the robot and the surrounding obstacles. When there is no obstacle blocking, the robot moves forward at full speed, and when it encounters an obstacle, it slows down, following the formula below:
其中，v为机器人移动速度，d1为机器人距障碍物距离，d2为紧急停止距离，β为速度比例系数，maxV为设定的机器人最大行驶速度；Among them, v is the moving speed of the robot, d 1 is the distance between the robot and the obstacle, d 2 is the emergency stop distance, β is the speed proportional coefficient, and maxV is the set maximum driving speed of the robot;
步骤103，偏移规则，在反应式导航中，移动机器人根据传感器信息进行局部路径规划，一般分为趋向目标行为与避障行为，若周围没有障碍物，机器人朝目标点以角度前进，前方有障碍物时，则需人为加入一个偏移噪声δ，机器人需无碰撞趋向目标，由此建立如下等式：Step 103, offset rule. In reactive navigation, the mobile robot performs local path planning according to the sensor information, which is generally divided into target behavior and obstacle avoidance behavior. If there are no obstacles around, the robot will move towards the target point by When advancing at an angle and there is an obstacle in front, it is necessary to artificially add an offset noise δ, and the robot needs to approach the target without collision, so the following equation is established:
Φn为移动机器人预瞄准方向，φn为n时刻角度，δn为n时刻偏移噪声；k为比例系数加入的偏移噪声大小，其值由模糊神经网络控制器根据机器人当前所处的环境确定，当时，器人朝向目标位置前进；当时，移动机器人将按照加入偏移角度后的目标方向前进。Φ n is the pre-aiming direction of the mobile robot, φ n is the angle at n time, δ n is the offset noise at n time; k is the size of the offset noise added by the proportional coefficient, and its value is determined by the fuzzy neural network controller according to the current position of the robot The environment is determined when When , the robot moves towards the target position; when , the mobile robot will move forward in the target direction after adding the offset angle.
本发明相比现有技术，具有以下有益效果：Compared with the prior art, the present invention has the following beneficial effects:
1.将神经网络与模糊控制的优点相结合，融合神经网络的自学习能力与模糊控制的模糊推理能力，减少逻辑推理工作量。1. Combine the advantages of neural network and fuzzy control, integrate the self-learning ability of neural network and the fuzzy reasoning ability of fuzzy control, and reduce the workload of logical reasoning.
2.改进型虚目标方法，采用简单的虚拟目标方法，摆脱机器人趋向目标运行中的陷阱状态。2. The improved virtual target method adopts a simple virtual target method to get rid of the trap state when the robot is running towards the target.
附图说明Description of drawings
图1为本发明的流程示意图；Fig. 1 is a schematic flow sheet of the present invention;
图2为ANFIS结构示意图。Figure 2 is a schematic diagram of the structure of ANFIS.
具体实施方式detailed description
下面结合附图和具体实施例，进一步阐明本发明，应理解这些实例仅用于说明本发明而不用于限制本发明的范围，在阅读了本发明之后，本领域技术人员对本发明的各种等价形式的修改均落于本申请所附权利要求所限定的范围。Below in conjunction with accompanying drawing and specific embodiment, further illustrate the present invention, should be understood that these examples are only for illustrating the present invention and are not intended to limit the scope of the present invention, after having read the present invention, those skilled in the art will understand various aspects of the present invention All modifications of the valence form fall within the scope defined by the appended claims of the present application.
一种基于ANFIS模糊神经网络的机器人路径规划方法，首先对移动机器人建立运动学模型；借助神经网络的自主学习功能和模糊理论的模糊推理能力，提出一种模糊神经网络的移动机器人导航的模糊神经网络控制器；其基于自适应模糊神经网络结构，构建Takagi-Sugeno型模糊推理系统并作为机器人局部反应控制的参考模型；将障碍物的距离和位置的相关信息作为模糊神经网络控制器的两个输入，模糊神经网络控制器实时输出机器人偏移角度和运行速度，通过模糊神经网络控制器在线调整移动机器人的偏移方向，使移动机器人能够无碰撞自动调整速度趋向目标。A robot path planning method based on the ANFIS fuzzy neural network. Firstly, a kinematics model is established for the mobile robot; with the help of the autonomous learning function of the neural network and the fuzzy reasoning ability of the fuzzy theory, a fuzzy neural network for mobile robot navigation is proposed. Network controller; based on the adaptive fuzzy neural network structure, a Takagi-Sugeno type fuzzy inference system is constructed and used as a reference model for robot local response control; the distance and position related information of obstacles are used as two of the fuzzy neural network controller Input, the fuzzy neural network controller outputs the robot's offset angle and running speed in real time, and adjusts the offset direction of the mobile robot online through the fuzzy neural network controller, so that the mobile robot can automatically adjust the speed to the target without collision.
通过模糊神经网络控制器输出值表示机器人移动角度和运行速度，越靠近障碍物时输出角度绝对值越大，速度绝对值越小；当前方没有障碍物时，机器人沿预设定方向前进；当前方有一个障碍物时，机器人逐渐接近障碍物，在一定范围内实时改变偏移角度和速度，使机器人缓慢绕离障碍物驶向目标；当前方有两个及其以上障碍物时，移动机器人在行进过程中对虚拟目标进行实时调整，即机器人沿所识别的最后一个障碍物前进并避开除此之外所有障碍物，选择一条远离障碍的最优路径趋向目标。The output value of the fuzzy neural network controller represents the moving angle and running speed of the robot. The closer the obstacle is, the larger the absolute value of the output angle is, and the smaller the absolute value of the speed is; when there is no obstacle in front, the robot moves forward in a preset direction; When there is an obstacle in front, the robot gradually approaches the obstacle, and changes the offset angle and speed in real time within a certain range, so that the robot slowly moves away from the obstacle to the target; when there are two or more obstacles in front, the mobile robot The virtual target is adjusted in real time during the traveling process, that is, the robot advances along the last obstacle identified and avoids all other obstacles, and chooses an optimal path away from the obstacle to approach the target.
模糊神经网络控制器利用LMS算法和最小二乘法来完成输入/输出数据对的建模，使得Takagi-Sugeno型模糊推理系统能模拟出希望或实际的输入/输出关系。模糊神经网络控制器在学习时，根据系统实际输出值与期望输出值计算出学习误差，再通过LMS算法对系统的偏移角度和运行速度进行调整。The fuzzy neural network controller uses the LMS algorithm and the least square method to complete the modeling of the input/output data pair, so that the Takagi-Sugeno type fuzzy inference system can simulate the desired or actual input/output relationship. When the fuzzy neural network controller is learning, it calculates the learning error according to the actual output value and the expected output value of the system, and then adjusts the offset angle and running speed of the system through the LMS algorithm.
针对位置环境下移动机器人导航实际问题，构建神经网络控制器，将障碍物的距离和位置的相关信息作为控制器的两个输入，机器人偏移角度和运行速度作为输出，实现了局部路径规划，并结合虚拟子目标的方法，能够增强系统解决传统反应导航问题中陷阱状态下的路径复杂和路径冗余问题。解决了传统反应导航问题中路径复杂和路径冗余问题，规划处一条逃离陷阱状态无碰撞趋向目标的最优路径。Aiming at the actual problem of mobile robot navigation in the location environment, a neural network controller is constructed, and the distance and position related information of obstacles are used as two inputs of the controller, and the robot's offset angle and running speed are used as outputs to realize local path planning. And combined with the method of virtual sub-goal, the system can be enhanced to solve the problem of path complexity and path redundancy in the trap state in the traditional reaction navigation problem. It solves the problem of complex path and path redundancy in the traditional reaction navigation problem, and plans an optimal path to escape from the trap state and approach the goal without collision.
1.通过机器人周围的传感器测量障碍物距离，并对机器人的位置，速度进行建模并建立避障规则。1. Measure the obstacle distance through the sensors around the robot, model the position and speed of the robot and establish obstacle avoidance rules.
(1)移动机器人通过本体携带的测距传感器测量障碍物的距离。机器人当前坐标为(xr,yr),目标点坐标为(xt,yt)，E是机器人当前位置(xr,yr)到目标点(xt,yt)的矢量，其模长和向量角表示为(1) The mobile robot measures the distance of obstacles through the ranging sensor carried by the body. The current coordinates of the robot are (x r , y r ), the coordinates of the target point are (x t , y t ), E is the vector from the current position of the robot (x r , y r ) to the target point (x t , y t ), and its The modulus length and vector angle are expressed as
En为机器人在目标距离势场中的势能；当前机器人与目标点的夹角，根据机器人当前位置不断修正，始终指向目标位置；下标n表示具体时刻。E n is the potential energy of the robot in the target distance potential field; The angle between the current robot and the target point is constantly corrected according to the current position of the robot, and always points to the target position; the subscript n indicates the specific moment.
(2)速度模型(2) Velocity model
移动机器人在导航任务中的速度由机器人与周围障碍物之间距离决定。当无障碍物阻挡时，机器人全速前进，当遇到障碍物时减速行驶。遵循以下公式：The speed of a mobile robot in a navigation task is determined by the distance between the robot and surrounding obstacles. When there is no obstacle blocking, the robot moves forward at full speed, and slows down when encountering an obstacle. Follow the formula below:
v为机器人移动速度；d1为机器人距障碍物距离；d2为紧急停止距离；β速度比例系数；maxV为设定的机器人最大行驶速度。v is the moving speed of the robot; d 1 is the distance between the robot and the obstacle; d 2 is the emergency stop distance; β speed proportional coefficient; maxV is the set maximum driving speed of the robot.
(3)偏移规则(3) Offset rules
在反应式导航中，移动机器人根据传感器信息进行局部路径规划。一般分为趋向目标行为与避障行为。若围没有障碍物，器人朝目标点以角度前进，前方有障碍物时，则需人为加入一个偏移噪声δ，机器人需无碰撞趋向目标，由此建立如下等式In reactive navigation, mobile robots perform local path planning based on sensor information. It is generally divided into goal-oriented behavior and obstacle avoidance behavior. If there are no obstacles around, the robot moves towards the target point When advancing at an angle and there is an obstacle in front, it is necessary to artificially add an offset noise δ, and the robot needs to approach the target without collision, thus establishing the following equation
为移动机器人预瞄准方向；k为比例系数加入的偏移噪声大小，其值由模糊神经网络控制器根据机器人当前所处的环境确定.当时，器人朝向目标位置前进；当时，移动机器人将按照加入偏移角度后的目标方向前进。is the pre-aiming direction of the mobile robot; k is the size of the offset noise added by the proportional coefficient, and its value is determined by the fuzzy neural network controller according to the current environment of the robot. When When , the robot moves towards the target position; when , the mobile robot will move forward in the target direction after adding the offset angle.
2.基于自适应模糊神经网络ANFIS网络，构建Takagi-Sugeno型模糊推理系统，提出神经网络控制器。2. Based on the adaptive fuzzy neural network ANFIS network, a Takagi-Sugeno type fuzzy inference system is constructed, and a neural network controller is proposed.
将障碍物的距离和位置的相关信息作为控制器的两个输入，机器人偏移角度和运行速度作为输出。模糊神经网络系统利用LMS算法和最小二乘法来完成输入/输出数据对的建模，使得Takagi-Sugeno型模糊推理系统能模拟出希望或实际的输入/输出关系。模糊神经系统在学习时，根据系统实际输出值与期望输出值可以计算出学习误差，再通过LMS算法对系统的偏移角度和运行速度进行调整。The information about the distance and position of the obstacle is used as the two inputs of the controller, and the robot's offset angle and running speed are taken as the output. The fuzzy neural network system uses the LMS algorithm and the least square method to complete the modeling of the input/output data pair, so that the Takagi-Sugeno type fuzzy reasoning system can simulate the desired or actual input/output relationship. When the fuzzy nervous system is learning, the learning error can be calculated according to the actual output value and the expected output value of the system, and then the offset angle and running speed of the system can be adjusted through the LMS algorithm.
利用神经网络引入学习机制，为模糊控制器自动提取模糊规则及模糊隶属函数，使整个系统成为模糊神经网络系统。其样本数据是基于实际训练的数据，采用的自适应模糊神经网络的ANFIS网络，构建Takagi-Sugeno型模糊推理系统。The neural network is used to introduce the learning mechanism, and the fuzzy rules and fuzzy membership functions are automatically extracted for the fuzzy controller, so that the whole system becomes a fuzzy neural network system. The sample data is based on the actual training data, and the ANFIS network of adaptive fuzzy neural network is used to construct the Takagi-Sugeno type fuzzy reasoning system.
典型ANFIS结构，如图2所示，其中，x1,x2是系统的输入，y是推理系统的输入，均可提供据对；网络同一层的每个节点具有相似的功能，用O1+i表示第一层第i个节点的输出，依此类推。A typical ANFIS structure is shown in Figure 2, where x 1 and x 2 are the inputs of the system, and y is the input of the reasoning system, both of which can provide data pairs; each node in the same layer of the network has similar functions, and O 1 +i means the output of the i-th node in the first layer, and so on.
第一层：本层节点将输入信号模糊化The first layer: the node of this layer fuzzifies the input signal
O1+i＝μAi(xi),i＝1,2 (5)O 1+i ＝μA i ( xi ),i=1,2 (5)
Oi+j＝μBj-2(x2),j＝3,4 (6)O i+j =μB j-2 (x 2 ),j=3,4 (6)
其中，Ai或Bj-2。是模糊集，如“多”，“少”等；μAi(xi)是模糊集的隶属函数。Among them, A i or B j-2 . is a fuzzy set, such as "more", "less" and so on; μA i ( xi ) is the membership function of the fuzzy set.
第二层：本层节点用于计算各条规则的适用度wi，即：将各输人信号的隶属度相乘，并将乘积作为本规则适用度。The second layer: nodes in this layer are used to calculate the applicability w i of each rule, namely: multiply the membership degrees of each input signal, and take the product as the applicability of this rule.
O2+i＝wi＝μAi(x1)μBi(x2),i＝1,2 (7)O 2+i ＝w i ＝μA i (x 1 )μB i (x 2 ),i=1,2 (7)
第三层：本层节点进行各条规则适用度的归一化计算，即：计算第i条规则的与全部规则适用The third layer: the node in this layer performs the normalized calculation of the applicability of each rule, that is, calculates the i-th rule and applies to all rules
O3,i＝w1'＝wi/(w1+w2),i＝1,2 (8)O 3,i =w 1 '=w i /(w 1 +w 2 ),i=1,2 (8)
第四层：本层节点用于计算各条规则的输出The fourth layer: the nodes in this layer are used to calculate the output of each rule
Ok,i＝wi'fi＝wi'(pixi+qix2+ri),i＝1,2 (9) Ok,i ＝w i 'f i ＝w i '(p i x i +q i x 2 +r i ),i=1,2 (9)
其中，为Sugeno型模糊系统的后项(结论)输出函数，当其为线性函数则称为“一阶系统”；若为常量则称为“0阶系统”。Among them, it is the postterm (conclusion) output function of the Sugeno type fuzzy system. When it is a linear function, it is called a "first-order system"; if it is a constant, it is called a "0-order system".
第五层：本层为单节点，用于计算系统的总输出Fifth layer: This layer is a single node, used to calculate the total output of the system
本系统常采用的是误差反传算法或是与最小二乘相结合的混合算法来训练相关参数，使得系统能够很好地模拟给定的样本数据。自适应神经模糊推理系统最大的特点就是，该系统是基于数据的建模方法。This system often uses the error backpropagation algorithm or the hybrid algorithm combined with the least squares to train the relevant parameters, so that the system can simulate the given sample data well. The biggest feature of the adaptive neuro-fuzzy reasoning system is that the system is a data-based modeling method.
模糊神经网络系统利用LMS算法和最小二乘法来完成输入/输出数据对的建模。使得生成出来的Takagi-Sugeno型模糊推理系统能模拟出希望或是实际的输入/输出关系。模糊神经系统在学习时，根据系统实际输出值与期望输出值可以计算出学习误差，再通过LMS算法对系统参数进行调整。The fuzzy neural network system utilizes the LMS algorithm and the least square method to complete the modeling of the input/output data pair. The generated Takagi-Sugeno type fuzzy reasoning system can simulate the desired or actual input/output relationship. When the fuzzy nervous system is learning, it can calculate the learning error according to the actual output value and the expected output value of the system, and then adjust the system parameters through the LMS algorithm.
3.采用虚拟目标方法进行路径规划3. Using virtual target method for path planning
采用改进虚拟目标方法，选择机器人能够逃离陷阱状态的最优路径，通过自适应模糊神经网络控制器实时输出偏移角度和运行速度，在线调整移动机器人的前进方向，使移动机器人能够无碰撞自动调整速度趋向目标。The improved virtual target method is used to select the optimal path for the robot to escape from the trap state, and the real-time output of the offset angle and running speed through the adaptive fuzzy neural network controller is used to adjust the forward direction of the mobile robot online, so that the mobile robot can automatically adjust without collision Speed towards target.
通过模糊神经网络系统输出值表示机器人移动角度和运行速度，越靠近障碍物时输出角度绝对值越大，速度绝对值越小。当前方没有障碍物时，机器人沿预设定方向前进；当前方有一个障碍物时，机器人逐渐接近障碍物，在一定范围内实时改变偏移角度和速度，使机器人缓慢绕离障碍物驶向目标；当前方有两个及其以上障碍物时，为避免所提出的虚拟目标中路径冗余的复杂问题(陷入陷阱状态)，移动机器人在行进过程中需要对虚拟目标进行实时调整，即机器人沿所识别的最后一个障碍物前进并避开除此之外所有障碍物，选择一条远离障碍的最优路径趋向目标，最终完成对目标点的导航。The output value of the fuzzy neural network system represents the moving angle and running speed of the robot. The closer to the obstacle, the larger the absolute value of the output angle is, and the smaller the absolute value of the speed is. When there is no obstacle in front, the robot moves forward in the preset direction; when there is an obstacle in front, the robot gradually approaches the obstacle, and changes the offset angle and speed in real time within a certain range, so that the robot slowly drives away from the obstacle. target; when there are two or more obstacles in front, in order to avoid the complex problem of path redundancy in the proposed virtual target (falling into a trap state), the mobile robot needs to adjust the virtual target in real time during the travel process, that is, the robot Advance along the last identified obstacle and avoid all other obstacles, choose an optimal path away from the obstacle to approach the target, and finally complete the navigation to the target point.
以上所述仅是本发明的优选实施方式，应当指出：对于本技术领域的普通技术人员来说，在不脱离本发明原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也应视为本发明的保护范围。The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications are also possible. It should be regarded as the protection scope of the present invention.
Claims (5)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710429378.7A CN107168324B (en) | 2017-06-08 | 2017-06-08 | A Robot Path Planning Method Based on ANFIS Fuzzy Neural Network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710429378.7A CN107168324B (en) | 2017-06-08 | 2017-06-08 | A Robot Path Planning Method Based on ANFIS Fuzzy Neural Network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107168324A true CN107168324A (en) | 2017-09-15 |
| CN107168324B CN107168324B (en) | 2020-06-05 |
Family
ID=59824693
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710429378.7A Expired - Fee Related CN107168324B (en) | 2017-06-08 | 2017-06-08 | A Robot Path Planning Method Based on ANFIS Fuzzy Neural Network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107168324B (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107727100A (en) * | 2017-10-16 | 2018-02-23 | 广东智爱机器人科技有限公司 | A kind of Mobile Robotics Navigation method for planning track based on apart from type fuzzy reasoning |
| CN108053067A (en) * | 2017-12-12 | 2018-05-18 | 深圳市易成自动驾驶技术有限公司 | Planing method, device and the computer readable storage medium of optimal path |
| CN109242294A (en) * | 2018-08-29 | 2019-01-18 | 国网河南省电力公司电力科学研究院 | Improve the power communication performance method for early warning and device of fuzzy neural network |
| CN109597404A (en) * | 2017-09-30 | 2019-04-09 | 徐工集团工程机械股份有限公司 | Road roller and its controller, control method and system |
| CN109782757A (en) * | 2018-12-30 | 2019-05-21 | 芜湖哈特机器人产业技术研究院有限公司 | A kind of path dispatching method of more AGV systems based on subsection scheduling |
| CN110196588A (en) * | 2019-03-28 | 2019-09-03 | 陕西理工大学 | Method for planning path for mobile robot based on networks decomposition |
| CN110262512A (en) * | 2019-07-12 | 2019-09-20 | 北京机械设备研究所 | A kind of mobile robot is detached from the barrier-avoiding method and system of U-shaped obstacle trap |
| CN110426950A (en) * | 2019-07-06 | 2019-11-08 | 大国重器自动化设备(山东)股份有限公司 | A kind of intelligent robot based on fuzzy logic |
| CN110672101A (en) * | 2019-09-20 | 2020-01-10 | 北京百度网讯科技有限公司 | Navigation model training method and device, electronic equipment and storage medium |
| CN111443603A (en) * | 2020-03-31 | 2020-07-24 | 东华大学 | Robot sharing control method based on self-adaptive fuzzy neural network system |
| CN111506104A (en) * | 2020-04-03 | 2020-08-07 | 北京邮电大学 | Method and device for planning position of unmanned aerial vehicle |
| CN112212867A (en) * | 2020-10-19 | 2021-01-12 | 中国科学技术大学 | Robot self-positioning and navigation method and system |
| CN112513563A (en) * | 2018-08-31 | 2021-03-16 | 株式会社小松制作所 | Work machine transported object specifying device, work machine transported object specifying method, completion model production method, and learning dataset |
| CN112882461A (en) * | 2019-11-29 | 2021-06-01 | 新东工业株式会社 | Robot, control method, and storage medium |
| CN113190013A (en) * | 2018-08-31 | 2021-07-30 | 创新先进技术有限公司 | Method and device for controlling terminal movement |
| CN113867366A (en) * | 2021-11-02 | 2021-12-31 | 福建省海峡智汇科技有限公司 | Mobile robot control method based on adaptive network fuzzy |
| CN114326734A (en) * | 2021-12-29 | 2022-04-12 | 中原动力智能机器人有限公司 | Path planning method and device |
| TWI763990B (en) * | 2019-04-22 | 2022-05-11 | 第一商業銀行股份有限公司 | Appraisal method and system of buildings based on urban and rural attributes |
| CN115035396A (en) * | 2022-08-09 | 2022-09-09 | 北京东方通网信科技有限公司 | Robot sight distance path determining method and device |
| CN119200410A (en) * | 2024-11-25 | 2024-12-27 | 山东建筑大学 | Intelligent construction-oriented tower crane obstacle avoidance track planning and self-adaptive tracking control method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101650568A (en) * | 2009-09-04 | 2010-02-17 | 湖南大学 | Method for ensuring navigation safety of mobile robots in unknown environments |
| CN101819041A (en) * | 2010-04-16 | 2010-09-01 | 北京航空航天大学 | Self-evolution ANFIS and UKF combined GPS/MEMS-INS integrated positioning error dynamic forecasting method |
| CN101887271A (en) * | 2010-07-19 | 2010-11-17 | 东莞职业技术学院 | Path planning method of mobile robot |
| US20140188273A1 (en) * | 2012-12-31 | 2014-07-03 | King Fahd University Of Petroleum And Minerals | Control method for mobile parallel manipulators |
| CN106168810A (en) * | 2016-09-18 | 2016-11-30 | 中国空气动力研究与发展中心高速空气动力研究所 | A kind of unmanned plane during flying obstacle avoidance system based on RTK and method |
-
2017
- 2017-06-08 CN CN201710429378.7A patent/CN107168324B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101650568A (en) * | 2009-09-04 | 2010-02-17 | 湖南大学 | Method for ensuring navigation safety of mobile robots in unknown environments |
| CN101819041A (en) * | 2010-04-16 | 2010-09-01 | 北京航空航天大学 | Self-evolution ANFIS and UKF combined GPS/MEMS-INS integrated positioning error dynamic forecasting method |
| CN101887271A (en) * | 2010-07-19 | 2010-11-17 | 东莞职业技术学院 | Path planning method of mobile robot |
| US20140188273A1 (en) * | 2012-12-31 | 2014-07-03 | King Fahd University Of Petroleum And Minerals | Control method for mobile parallel manipulators |
| CN106168810A (en) * | 2016-09-18 | 2016-11-30 | 中国空气动力研究与发展中心高速空气动力研究所 | A kind of unmanned plane during flying obstacle avoidance system based on RTK and method |
Non-Patent Citations (5)
| Title |
|---|
| MOSHEN KUAI: "Research of Planetary Gear Fault Diagnosis Based on Permutation Entropy of CEEMDAN and ANFIS", 《SENSORS》 * |
| 付宜利 等: "基于模糊控制的自主机器人路径规划策略研究", 《机器人》 * |
| 宋颖丽 等: "基于ANFIS的移动机器人避障行为", 《山东理工大学学报（自然科学版）》 * |
| 张桥: "模糊神经网络信息融合方法在机器人避障中的应用", 《农业装备与车辆工程》 * |
| 李会来 等: "基于T-S型模糊神经网络的轮式机器人避障方法研究", 《计算机测量与控制》 * |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109597404A (en) * | 2017-09-30 | 2019-04-09 | 徐工集团工程机械股份有限公司 | Road roller and its controller, control method and system |
| CN107727100A (en) * | 2017-10-16 | 2018-02-23 | 广东智爱机器人科技有限公司 | A kind of Mobile Robotics Navigation method for planning track based on apart from type fuzzy reasoning |
| CN107727100B (en) * | 2017-10-16 | 2023-04-07 | 广东智爱机器人科技有限公司 | Mobile robot navigation track planning method based on distance type fuzzy reasoning |
| CN108053067A (en) * | 2017-12-12 | 2018-05-18 | 深圳市易成自动驾驶技术有限公司 | Planing method, device and the computer readable storage medium of optimal path |
| CN109242294A (en) * | 2018-08-29 | 2019-01-18 | 国网河南省电力公司电力科学研究院 | Improve the power communication performance method for early warning and device of fuzzy neural network |
| CN112513563A (en) * | 2018-08-31 | 2021-03-16 | 株式会社小松制作所 | Work machine transported object specifying device, work machine transported object specifying method, completion model production method, and learning dataset |
| CN113190013B (en) * | 2018-08-31 | 2023-06-27 | 创新先进技术有限公司 | Method and device for controlling movement of terminal |
| CN113190013A (en) * | 2018-08-31 | 2021-07-30 | 创新先进技术有限公司 | Method and device for controlling terminal movement |
| CN109782757A (en) * | 2018-12-30 | 2019-05-21 | 芜湖哈特机器人产业技术研究院有限公司 | A kind of path dispatching method of more AGV systems based on subsection scheduling |
| CN110196588A (en) * | 2019-03-28 | 2019-09-03 | 陕西理工大学 | Method for planning path for mobile robot based on networks decomposition |
| TWI763990B (en) * | 2019-04-22 | 2022-05-11 | 第一商業銀行股份有限公司 | Appraisal method and system of buildings based on urban and rural attributes |
| CN110426950A (en) * | 2019-07-06 | 2019-11-08 | 大国重器自动化设备(山东)股份有限公司 | A kind of intelligent robot based on fuzzy logic |
| CN110262512A (en) * | 2019-07-12 | 2019-09-20 | 北京机械设备研究所 | A kind of mobile robot is detached from the barrier-avoiding method and system of U-shaped obstacle trap |
| CN110262512B (en) * | 2019-07-12 | 2022-03-29 | 北京机械设备研究所 | Obstacle avoidance method and system for moving robot to separate from U-shaped obstacle trap |
| CN110672101A (en) * | 2019-09-20 | 2020-01-10 | 北京百度网讯科技有限公司 | Navigation model training method and device, electronic equipment and storage medium |
| CN110672101B (en) * | 2019-09-20 | 2021-09-28 | 北京百度网讯科技有限公司 | Navigation model training method and device, electronic equipment and storage medium |
| CN112882461A (en) * | 2019-11-29 | 2021-06-01 | 新东工业株式会社 | Robot, control method, and storage medium |
| CN111443603A (en) * | 2020-03-31 | 2020-07-24 | 东华大学 | Robot sharing control method based on self-adaptive fuzzy neural network system |
| CN111506104B (en) * | 2020-04-03 | 2021-10-01 | 北京邮电大学 | A method and device for planning the position of an unmanned aerial vehicle |
| CN111506104A (en) * | 2020-04-03 | 2020-08-07 | 北京邮电大学 | Method and device for planning position of unmanned aerial vehicle |
| CN112212867A (en) * | 2020-10-19 | 2021-01-12 | 中国科学技术大学 | Robot self-positioning and navigation method and system |
| CN112212867B (en) * | 2020-10-19 | 2024-05-28 | 中国科学技术大学 | A method and system for robot self-positioning and navigation |
| CN113867366A (en) * | 2021-11-02 | 2021-12-31 | 福建省海峡智汇科技有限公司 | Mobile robot control method based on adaptive network fuzzy |
| CN114326734A (en) * | 2021-12-29 | 2022-04-12 | 中原动力智能机器人有限公司 | Path planning method and device |
| CN114326734B (en) * | 2021-12-29 | 2024-03-08 | 中原动力智能机器人有限公司 | Path planning method and device |
| CN115035396A (en) * | 2022-08-09 | 2022-09-09 | 北京东方通网信科技有限公司 | Robot sight distance path determining method and device |
| CN119200410A (en) * | 2024-11-25 | 2024-12-27 | 山东建筑大学 | Intelligent construction-oriented tower crane obstacle avoidance track planning and self-adaptive tracking control method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107168324B (en) | 2020-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107168324B (en) | A Robot Path Planning Method Based on ANFIS Fuzzy Neural Network | |
| Pandey et al. | Autonomous mobile robot navigation between static and dynamic obstacles using multiple ANFIS architecture | |
| CN109375632B (en) | Real-time trajectory planning method for automatic driving vehicle | |
| Mohanty et al. | Navigation of autonomous mobile robot using adaptive network based fuzzy inference system | |
| CN106909144A (en) | For the unpiloted field obstacle-avoiding route planning of agricultural machinery and its control method | |
| CN104501816A (en) | Multi-unmanned aerial vehicle coordination and collision avoidance guide planning method | |
| Hart et al. | Graph neural networks and reinforcement learning for behavior generation in semantic environments | |
| CN110488835B (en) | An intelligent local path planning method for unmanned systems based on double back-propagation neural networks | |
| Bao et al. | A fuzzy behavior-based architecture for mobile robot navigation in unknown environments | |
| Al-Sagban et al. | Neural-based navigation of a differential-drive mobile robot | |
| Mohanty et al. | Path planning of an autonomous mobile robot using adaptive network based fuzzy controller | |
| Mohanty et al. | Path generation and obstacle avoidance of an autonomous mobile robot using intelligent hybrid controller | |
| Jajulwar et al. | Design of SLAM based adaptive fuzzy tracking controller for autonomous navigation system | |
| Tanha et al. | Control a mobile robot in Social environments by considering human as a moving obstacle | |
| CN113485323A (en) | Flexible formation method for cascaded multiple mobile robots | |
| Lei et al. | A fuzzy behaviours fusion algorithm for mobile robot real-time path planning in unknown environment | |
| Mohanty et al. | Navigation of an autonomous mobile robot using intelligent hybrid technique | |
| CN115586774A (en) | Mobile robot obstacle avoidance method based on improved dynamic window method | |
| CN107727100A (en) | A kind of Mobile Robotics Navigation method for planning track based on apart from type fuzzy reasoning | |
| Ai et al. | Neuro-fuzzy mobile robot navigation | |
| Boufera et al. | Fuzzy inference system optimization by evolutionary approach for mobile robot navigation | |
| Nguyen et al. | Social constraints-based socially aware navigation framework for mobile service robots | |
| Ma et al. | Fuzzy neural network-based real-time self-reaction of mobile robot in unknown environments | |
| Mohanty et al. | A new intelligent approach for mobile robot navigation | |
| Zhang et al. | Recent advances in robot trajectory planning in a dynamic environment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200605 Termination date: 20210608 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |