CN116880514B - Intelligent wheelchair control method, intelligent wheelchair and storage medium - Google Patents
Intelligent wheelchair control method, intelligent wheelchair and storage medium Download PDFInfo
- Publication number
- CN116880514B CN116880514B CN202311141989.3A CN202311141989A CN116880514B CN 116880514 B CN116880514 B CN 116880514B CN 202311141989 A CN202311141989 A CN 202311141989A CN 116880514 B CN116880514 B CN 116880514B
- Authority
- CN
- China
- Prior art keywords
- wheelchair
- obstacle
- path
- user
- preset
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000003860 storage Methods 0.000 title claims abstract description 13
- 230000007613 environmental effect Effects 0.000 claims description 52
- 230000008859 change Effects 0.000 claims description 25
- 238000009826 distribution Methods 0.000 claims description 24
- 230000033001 locomotion Effects 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 16
- 230000006399 behavior Effects 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 13
- 230000015654 memory Effects 0.000 claims description 9
- 238000004088 simulation Methods 0.000 claims description 7
- 230000007423 decrease Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 10
- 230000004044 response Effects 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 7
- 238000012790 confirmation Methods 0.000 description 7
- 230000004888 barrier function Effects 0.000 description 5
- 210000004556 brain Anatomy 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 5
- 230000003068 static effect Effects 0.000 description 5
- 230000006872 improvement Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 230000004424 eye movement Effects 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000004807 localization Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 208000006011 Stroke Diseases 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61G—TRANSPORT, PERSONAL CONVEYANCES, OR ACCOMMODATION SPECIALLY ADAPTED FOR PATIENTS OR DISABLED PERSONS; OPERATING TABLES OR CHAIRS; CHAIRS FOR DENTISTRY; FUNERAL DEVICES
- A61G5/00—Chairs or personal conveyances specially adapted for patients or disabled persons, e.g. wheelchairs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61G—TRANSPORT, PERSONAL CONVEYANCES, OR ACCOMMODATION SPECIALLY ADAPTED FOR PATIENTS OR DISABLED PERSONS; OPERATING TABLES OR CHAIRS; CHAIRS FOR DENTISTRY; FUNERAL DEVICES
- A61G2203/00—General characteristics of devices
- A61G2203/10—General characteristics of devices characterised by specific control means, e.g. for adjustment or steering
- A61G2203/18—General characteristics of devices characterised by specific control means, e.g. for adjustment or steering by patient's head, eyes, facial muscles or voice
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61G—TRANSPORT, PERSONAL CONVEYANCES, OR ACCOMMODATION SPECIALLY ADAPTED FOR PATIENTS OR DISABLED PERSONS; OPERATING TABLES OR CHAIRS; CHAIRS FOR DENTISTRY; FUNERAL DEVICES
- A61G2203/00—General characteristics of devices
- A61G2203/10—General characteristics of devices characterised by specific control means, e.g. for adjustment or steering
- A61G2203/22—General characteristics of devices characterised by specific control means, e.g. for adjustment or steering for automatically guiding movable devices, e.g. stretchers or wheelchairs in a hospital
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61G—TRANSPORT, PERSONAL CONVEYANCES, OR ACCOMMODATION SPECIALLY ADAPTED FOR PATIENTS OR DISABLED PERSONS; OPERATING TABLES OR CHAIRS; CHAIRS FOR DENTISTRY; FUNERAL DEVICES
- A61G2203/00—General characteristics of devices
- A61G2203/70—General characteristics of devices with special adaptations, e.g. for safety or comfort
- A61G2203/72—General characteristics of devices with special adaptations, e.g. for safety or comfort for collision prevention
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Animal Behavior & Ethology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses an intelligent wheelchair control method, an intelligent wheelchair and a storage medium, wherein obstacle information is obtained by detecting obstacles in an environment space, when the position of the wheelchair is judged to need obstacle avoidance and can avoid the obstacle according to the obstacle information, behavior prediction is carried out on dynamic obstacles, collision probability is calculated, an environment risk level is judged, and when the environment risk level is not high, automatic obstacle avoidance is carried out. The autonomous decision, autonomous navigation and intelligent obstacle avoidance of the wheelchair are realized, so that the intelligent wheelchair can be better adapted to various complex environments, the obstacle avoidance and stability are improved, the problem of frequent alarm is effectively reduced, and the use effect and safety of the intelligent wheelchair are improved.
Description
Technical Field
The invention relates to the technical field of intelligent control, in particular to an intelligent wheelchair control method, an intelligent wheelchair and a storage medium.
Background
The intelligent wheelchair realizes autonomous navigation mainly through machine vision, SLAM (Simultaneous Localization And Mapping: simultaneous localization and mapping) and path planning technologies.
The current intelligent wheelchair control method is based on static obstacles in a space environment to carry out path planning so as to realize an obstacle avoidance function, and only can better avoid the static obstacles, when dynamic obstacles (such as pedestrians or bicycles) move nearby the wheelchair, alarming or braking actions are adopted, and frequent execution of the actions can cause unsmooth wheelchair running in a complex environment, so that the user experience is poor.
Accordingly, there is a need for improvement and advancement in the art.
Disclosure of Invention
The invention mainly aims to provide an intelligent wheelchair control method, an intelligent wheelchair and a computer readable storage medium, which solve the problems of unsmooth running in a complex environment and poor user experience during wheelchair control.
To achieve the above object, a first aspect of the present invention provides an intelligent wheelchair control method, including:
planning a navigation path based on the selected destination, and controlling the wheelchair to move along the navigation path by adopting an automatic navigation mode;
in the automatic navigation process, detecting an obstacle in an environment space where the wheelchair is positioned to obtain obstacle information;
obtaining the risk type of the position of the wheelchair according to the positions of all the obstacles in the obstacle information and the navigation path;
When the risk type is obstacle avoidance and obstacle avoidance is possible:
performing behavior prediction on the dynamic obstacle in the obstacle to obtain the position distribution information of the dynamic obstacle in a preset time step;
calculating collision probability of each dynamic obstacle and the navigation path according to the position distribution information, and obtaining an environmental risk level according to all the collision probabilities;
and when the environmental risk level is high risk, controlling the wheelchair to move by adopting a manual control mode, otherwise, performing obstacle avoidance planning according to the obstacle information and the position distribution information to obtain a safety path, and controlling the wheelchair to move along the safety path within a preset time step so as to automatically avoid the obstacle.
Optionally, the risk type of obtaining the position of the wheelchair according to the positions of all the obstacles in the obstacle information and the navigation path includes:
calculating a first distance between the position of each obstacle and the center of the wheelchair and a second distance between each obstacle and the navigation path;
when any one of the first distances is smaller than a first distance threshold, setting the risk type as an unavoidable obstacle;
otherwise, when all the second distances are larger than a second distance threshold, setting the risk type as not needing obstacle avoidance; and when any one of the second distances is smaller than or equal to a second distance threshold value, setting the risk type to be obstacle avoidance and obstacle avoidance.
Optionally, the calculating, according to the position distribution information, a collision probability of each dynamic obstacle with the navigation path includes:
performing wheelchair movement simulation for preset times, and obtaining collision times of each dynamic obstacle and the navigation path according to the position distribution information;
and calculating the collision probability according to the category weight of each dynamic obstacle, the collision times, the preset times and the probability value in the position distribution information.
Optionally, the performing obstacle avoidance planning according to the obstacle information and the position distribution information to obtain a safety path includes:
calculating potential field values of all positions of the navigation path region;
taking the direction of the decline of the potential field value as the moving direction of the wheelchair to obtain a plurality of obstacle avoidance paths;
accumulating potential field values of all track points on each obstacle avoidance path to obtain a total potential field value of each obstacle avoidance path;
and setting the obstacle avoidance path with the minimum total potential field value as the safety path.
Optionally, when the environmental risk level is low risk, the collision early warning function is turned off, and the wheelchair is controlled to move along the safety path within a preset time step so as to automatically avoid the obstacle, including:
Calculating a first time window according to the collision probability, and controlling the wheelchair to move along the safety path in the first time window;
when the first time window is over, the collision probability of each dynamic obstacle and the navigation path is recalculated, and the environmental risk level is updated according to all the collision probabilities;
and when the updated environment risk level is low risk, controlling the wheelchair to move along the safety path within a preset second time window so as to prolong the automatic navigation time.
Optionally, when the wheelchair is controlled to move along the safety path within a preset second time window and an instruction of selecting the manual control mode by the user is received, the method further includes:
collecting a first electroencephalogram signal of a user in real time;
when the user is in a vigilance state according to the first electroencephalogram signal and a pre-established electroencephalogram signal baseline, after a preset time interval, acquiring a second electroencephalogram signal of the user in real time, and when the user is still in the vigilance state according to the second electroencephalogram signal and the pre-established electroencephalogram signal baseline, responding to a command of the user for confirming entering a manual control mode, controlling the wheelchair to move by adopting the manual control mode and starting a collision early warning function.
Optionally, when the wheelchair is controlled to move along the safety path within a preset second time window and an instruction of selecting the manual control mode by the user is received, the method further includes:
counting the use time of an automatic navigation mode in a preset time period;
when the using time is larger than a preset threshold, responding to a command of a user for confirming to enter a manual control mode, and controlling the wheelchair to move and start a collision early warning function in a preset third time window by adopting the manual control mode.
Optionally, in the process of automatically avoiding the obstacle, the method further comprises:
comparing the collision probability of the current preset time step with the collision probability of the previous preset time step to obtain an environmental risk change proportion;
and if the continuous preset number of environmental risk change ratios exceed a preset threshold, reducing the duration of the first time window and the second time window, and sending an alarm prompt.
A second aspect of the present invention provides a smart wheelchair comprising a memory, a processor, and a smart wheelchair control program stored on the memory and executable on the processor, the smart wheelchair control program implementing the steps of any one of the smart wheelchair control methods when executed by the processor.
A third aspect of the present invention provides a computer-readable storage medium having stored thereon an intelligent wheelchair control program that, when executed by a processor, implements the steps of any one of the above-described intelligent wheelchair control methods.
From the above, the invention obtains the obstacle information by detecting the obstacle in the environment space, when the position of the wheelchair is judged to need obstacle avoidance and can avoid the obstacle according to the obstacle information, the behavior prediction is carried out on the dynamic obstacle, the collision probability is calculated, the environment risk level is judged, and when the environment risk level is not high risk, the automatic obstacle avoidance is carried out. The wheelchair can realize autonomous decision, autonomous navigation and intelligent obstacle avoidance. The intelligent wheelchair can be better adapted to various complex environments, the obstacle avoidance and stability performances of the intelligent wheelchair are improved, the problem of frequent alarm is effectively reduced, and the use effect and the safety of the intelligent wheelchair are improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a control method of an intelligent wheelchair according to an embodiment of the present invention;
FIG. 2 is a flow chart of a secure path acquisition according to an embodiment of the present invention;
FIG. 3 is a schematic flow chart of an automatic obstacle avoidance according to an embodiment of the present invention;
FIG. 4 is a flow chart of a user when receiving an instruction to select a manual control mode;
FIG. 5 is a flow chart of a method for selecting a manual control mode when an instruction is received from a user and the environmental risk level is low risk;
FIG. 6 is a flow chart illustrating the environment risk change during the wheelchair movement according to the embodiment of the present invention;
fig. 7 is a schematic block diagram of an internal structure of an intelligent wheelchair according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made more fully hereinafter with reference to the accompanying drawings, in which embodiments of the invention are shown, it being evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways other than those described herein, and persons skilled in the art will readily appreciate that the present invention is not limited to the specific embodiments disclosed below.
When navigating in complicated indoor and outdoor environments (such as places with more people and more obstacles, such as parks, urban sets, supermarkets, hospitals and the like), the existing wheelchair control method directly alarms if the possibility and necessity of obstacle avoidance are not fully judged, so that the real dangerous situation cannot be judged due to too frequent alarm, even if dynamic obstacles with risks are detected, the wheelchair is directly braked, the wheelchair is not continuously driven, and the user experience is poor.
Aiming at the problems, the invention provides an intelligent wheelchair control method, which is used for calculating the collision probability of a dynamic obstacle when the position of the wheelchair is determined to be obstacle-avoiding and obstacle-avoiding, determining the environmental risk according to the collision probability, adopting a manual control mode when the environmental risk is high risk, otherwise, performing obstacle-avoiding planning to obtain a safety path, and then controlling the wheelchair to move along the safety path.
Method embodiment
The embodiment of the invention provides an intelligent wheelchair control method which is deployed on a control chip of a wheelchair and is used for accurately controlling the intelligent wheelchair in a complex environment, so that the control safety can be realized, and the control fluency can be improved.
As shown in fig. 1, the present embodiment specifically includes the following steps:
Step S100: planning a navigation path based on the selected destination, and controlling the wheelchair to move along the navigation path by adopting an automatic navigation mode;
step S200: in the automatic navigation process, detecting an obstacle in an environment space where the wheelchair is positioned to obtain obstacle information;
and automatically planning a navigation path from the position of the wheelchair to the optimal destination position according to the destination position selected by the user on the map interface, and controlling the wheelchair to automatically navigate to the destination according to the navigation path.
When planning a path, useThe algorithm calculates the optimal navigation path. Specifically, a series of connected track points are first generated in a navigation map according to the location of the wheelchair and the destination location selected by the user. And then evaluating the distance between the track points, the steering difficulty and the obstacle condition, and selecting the optimal navigation path. And the optimal navigation path avoids all known static obstacles and takes into account wheelchair size, steering radius, etc. The static obstacle can be various obstacles on the navigation map, such as a fence of a flower nursery; it may be a pedestrian or a vehicle in a stationary state.
When the wheelchair approaches to the next track point according to the data of the environment sensor, the wheelchair starts to decelerate, and the environment sensor scans the environment around the track point, and after safety is judged, steering or direction adjustment is performed, and the wheelchair is driven to the next track point. In the process of driving from one track point to the next track point, the environment sensor continuously scans the image information of the surrounding environment of the position of the wheelchair to obtain barrier information, and the barrier information is matched with map data to position the position and face direction of the wheelchair in real time. When the last track point is the destination track point, the wheelchair is controlled to slowly travel to the vicinity of the destination and scan the surrounding environment. After confirming the safety, the wheelchair is stopped, and a prompt tone is sent to prompt the user that the destination of the user has arrived. The position setting of the track points on the navigation path is not limited, and the track points can be flexibly set according to the actual space environment.
In the embodiment, the environment sensor is a camera and a three-dimensional laser radar which are arranged in front of the wheelchair, and the measuring range of the environment sensor covers the traveling direction by 6-8 meters. The environment sensor scans the surrounding environment in real time at a fixed frequency, such as 10Hz, and acquires the position and image information of the wheelchair, wherein the image resolution is 0.5-2 meters.
The obstacle information mainly includes the position and category of the obstacle (such as pedestrians, bicycles, cars, piled up sundries, etc.). After the environmental image is obtained, obstacle information is obtained by inputting the environmental image into an obstacle recognition model. It is necessary to first train an obstacle recognition model: and collecting an environment image data set, acquiring the position and the category of the obstacle in the image by adopting a manual labeling method, and constructing a data set D. The obstacle recognition model M is trained using the data set D as training data using a target detection algorithm (e.g., YOLOv 3). After training, the obstacle recognition model M can recognize the obstacle position (x i ,y i ) Category c i 。
In one example, the destination is selected by voice. Specifically, a user voice instruction is recognized through a voice recognition module: when the user sends out a voice instruction, such as 'going to a property management department' or 'going to 12 buildings', the voice recognition module recognizes the voice instruction of the user and analyzes the destination name. And searching in a map database by taking the analyzed destination name as a search keyword to obtain position information such as coordinates, directions and the like. If multiple destinations of the same name are searched, the user is asked which one is selected, and the user selects a specific destination again by voice instruction. After the destination position is obtained, an optimal navigation path is automatically planned, and the optimal navigation path is informed to a user in a voice mode, and the user waits for confirmation. After the voice confirmation of the user, the wheelchair is controlled to navigate to the destination according to the navigation path.
In another example, the destination is selected by an electroencephalogram signal. Specifically, the brain control module is used for identifying the brain electrical signal instruction of the user. The user concentrates on imagining to navigate to a certain destination, and acquires the electroencephalogram signals through the electroencephalogram head ring, identifies the electroencephalogram signals of the user, analyzes the destination name and obtains the destination position. If the signal is not clear, a plurality of options are presented for the user to confirm secondarily. After the destination information is obtained, an optimal navigation path is planned, the navigation path is presented to the user in a visual and voice mode, and a confirmation instruction of the user is waited. After the user sends out the confirmation instruction again through brain control or voice, the wheelchair is controlled to navigate to the destination according to the navigation path.
Step S300: obtaining the risk type of the position of the wheelchair according to the positions of all the obstacles in the obstacle information and the navigation path;
the risk type is used for measuring the safety of the position of the wheelchair and is mainly divided into the following types: the method does not need to avoid the obstacle, can not avoid the obstacle, and can avoid the obstacle. The wheelchair is safe in position without obstacle avoidance, and collision cannot occur; the unavoidable obstacle means that the wheelchair is very dangerous in the position, the possibility of collision is very high, and the wheelchair is not easy to avoid; the obstacle avoidance is needed, and the obstacle avoidance refers to the risk of collision at the position where the wheel chair is located, but the obstacle avoidance treatment can be used for avoiding the collision.
Specifically, the positions (x i ,y i ) And category c i . Then calculate the wheelchair center (x in image I c ,y c ) With each obstacle position (x i ,y i ) The Euclidean distance between the two is used for obtaining a first distance, and the specific expression is as follows:. If the first distance dsq of any obstacle is smaller than the first distance threshold d0 (0.5 m in this embodiment), the wheelchair is determined to be in an emergency position, and the risk type is unavoidable.
When the first distance dsq of all the obstacles is greater than or equal to the first distance threshold d0, each obstacle position (x i ,y i ) And a second distance is obtained from the vertical distance between the navigation path and the navigation path. If the second distance of all obstacles is greater than the second distance threshold ths (also calledFor a safe distance threshold, in generalMeter), determining the position of the wheelchair as a safe position, wherein the risk type is that obstacle avoidance is not needed; when the second distance of any obstacle is smaller than or equal to the second distance threshold, the position of the wheelchair is judged to be a dangerous position, and the risk type is obstacle avoidance and obstacle avoidance.
When the position of the wheelchair is a safe position, controlling the wheelchair to move in an automatic navigation mode; when the position of the wheelchair is an emergency position, decelerating or emergency braking is performed and a manual control mode is required to be adopted; when the wheelchair is in a dangerous position, namely, the risk type is obstacle avoidance and obstacle avoidance, the following steps S400-S600 are executed, and the wheelchair is automatically obstacle-avoided through operations such as obstacle behavior prediction, collision prediction, obstacle avoidance planning, alarming and the like. In the process of automatic obstacle avoidance, if an obstacle which cannot avoid the obstacle exists between any two track points or the obstacle avoidance fails, the wheelchair is stopped from moving, the automatic navigation mode is exited, an alarm is sent to prompt a user, and then the step S100 is returned to plan a navigation path again.
Step S400: predicting the behavior of the dynamic obstacle in the obstacle to obtain the position distribution information of the dynamic obstacle in a preset time step;
unlike static obstacles, dynamic obstacles change their positions with time, so that when automatically avoiding obstacles, it is necessary to predict the behavior of dynamic obstacles.
Specifically, a large number of videos of real scenes (such as communities and parks) are collected, the motion trail and parameters of various dynamic obstacles are extracted, and a behavior data set is constructed. An LSTM (Long Short Term Memory: long-short-term memory network) neural network is selected to construct a behavior prediction model, and the behavior prediction model is trained by using a behavior data set, so that the trained behavior prediction model can predict the motion state and the position of a dynamic barrier within a preset time step (generally 5-10 seconds). After training, selecting characteristics such as category, centroid position, size, speed, acceleration and direction from the obstacle information as motion trail and parameters, inputting the motion trail and parameters into a behavior prediction model, predicting future motion parameters of the dynamic obstacle in a preset time step by the behavior prediction model, and mapping the future motion parameters to the current environment to obtain the position distribution information of the dynamic obstacle.
In one example, the position distribution information is also patterned to obtain a probability prediction graph of the position of the dynamic obstacle, so that a more visual display effect can be obtained when collision early warning display is performed.
Step S500: calculating collision probability of each dynamic obstacle and a navigation path according to the position distribution information, and obtaining an environmental risk level according to all the collision probabilities;
and selecting a collision probability calculation method, such as a Monte Carlo algorithm, and performing wheelchair movement simulation for preset times, wherein the simulation time length is a preset movement step length. And simulating an obstacle moving track according to the position distribution information of the obstacles, counting the collision times n, n/(total simulation times) of each dynamic obstacle and the navigation path to obtain the collision probability of each dynamic obstacle and the navigation path, accumulating the collision probabilities of all the dynamic obstacles to obtain the collision probability of the navigation path, and comparing the collision probability of the navigation path with a preset threshold value point to judge the environmental risk level.
In order to make the collision probability of the navigation path more accurate, the embodiment also considers that the influence results of dynamic obstacles of different categories are different, for example, the weight of the collision probability of a person or a bicycle should be higher, and the wheelchair moving track within the preset time step is divided into a plurality of sections of sub-paths. When the collision probability of the dynamic obstacle is calculated, comprehensive weighted calculation is carried out according to the category weight, the collision times, the total simulation times and the probability value in the position distribution information of the dynamic obstacle, so as to obtain the collision probability of the more accurate navigation path. Specifically: firstly, dynamic obstacle information { O } obtained according to a behavior prediction model i Obtaining category c of each obstacle i Position (x) i ,y i ) Velocity v i Direction of movement theta i Equal data, obtaining track information { P } of each segment of sub-path in navigation path j }(x j ,y j ) Direction of movement theta j . Assuming a moving step of the dynamic obstacle of(the specific value is determined by weighing the calculation cost and the accuracy, and generally takes 0.1-0.5 seconds) according to the obstacle category c i The weight of the obstacle is determined. The weight of a person or bicycle can be generally set to be large. Simultaneously calculating the +.>Position prediction probability distribution in time step +.>，/>The more concentrated, the more accurate the prediction. To be used forFor collision detection position, each obstacle O is calculated i And sub-path P j Distance between->. If->Record dynamic obstacle O i And sub-path P j A collision occurs. Wherein R is i Is a dynamic obstacle O i Collision detection radius, R j Is the sub-path width. Repeating simulation for N times on each segment of sub-path to obtain sub-path P j With each obstacle O i Is +.>. Consider obstacle O i Category weight w of (2) ci Probability +.>The collision probability calculation formula of each dynamic obstacle and the sub-path is as follows: />. Accumulating collision probability of all dynamic barriers in the sub-path j to obtain collision probability +. >Where i=1 to n, n is the total number of dynamic obstacles. The maximum value of the collision probabilities of all the sub-paths is the collision probability P of the navigation path. Probability of collision P when navigating a path<0.3, judging that the environmental risk level is low risk and the value is 1; when P is more than or equal to 0.3<0.5, judging that the environmental risk level is stroke risk, and the value is 2; when P is more than or equal to 0.5, the environmental risk level is judged to be high risk, and the value is 3.
Step S600: when the environmental risk level is high risk, a manual control mode is adopted to control the wheelchair to move, otherwise, obstacle avoidance planning is carried out according to obstacle information and position distribution information, a safety path is obtained, and the wheelchair is controlled to move along the safety path within a preset time step so as to avoid the obstacle automatically.
When the environmental risk level is high, the probability of collision caused by dynamic obstacles in the space environment is high, and the wheelchair is controlled to move in a manual control mode in order to ensure the use safety. After the manual control is finished, the collision probability P of the navigation path is recalculated, if the P value is reduced and becomes 0.3-0.5, the switching operation is prompted and the countdown is displayed, the user does not operate and directly enters an automatic navigation mode, the situation that the automatic navigation mode can be restored only by manual confirmation after the manual control mode is finished is avoided, and the interaction efficiency is improved; if the duration P is more than or equal to 0.5, the manual control state is maintained.
When the environmental risk level is low risk or medium risk, obstacle avoidance planning is performed according to the obstacle information and the position distribution information, so as to obtain a safety path, wherein the safety path should avoid dynamic obstacles and a prediction area and be as close to a normal path as possible. And controlling the wheelchair to move along the safety path within a preset time step so as to realize automatic obstacle avoidance.
In this embodiment, during obstacle avoidance planning, the environmental image, the obstacle information and the position distribution information are input into an obstacle avoidance algorithm, such as an artificial potential field method, to obtain a safe path. As shown in fig. 2, the specific steps include:
step S610: calculating potential field values of all positions of the navigation path region;
step S620: taking the direction of the decline of the potential field value as the moving direction of the wheelchair to obtain a plurality of obstacle avoidance paths;
step S630: accumulating potential field values of all track points on each obstacle avoidance path to obtain a total potential field value of each obstacle avoidance path;
step S640: and setting the obstacle avoidance path with the minimum total potential field value as a safety path.
Specifically, the current position is taken as a starting point, the position reached by the wheelchair moving for a preset time step is taken as an ending point, and the area between the starting point and the ending point is a navigation path area, and the wheelchair comprises not only a navigation path between the starting point and the ending point, but also an area within a certain range around the navigation path.
The obstacle information includes the position (x i ,y i ) Dimension d i When obstacle avoidance planning is performed, if the distance d between the dynamic obstacle and the navigation path is smaller than the preset safety distance r s And judging that the wheelchair collides with the dynamic obstacle. Preferably, the safety distance r is preset s Is not a fixed value, and is 1-1.5 times of the size of the dynamic barrier.
The potential field value U (x, y) of each position (x, y) in the navigation path region is calculated according to the potential field function, and the specific calculation expression is as follows:wherein k is 1 Is constant, d 0 The range of action of the potential field function is determined. When->When U (x, y) gradually approaches 0.
After the potential field value U of each position is calculated, according to the principle that the direction with smaller potential field value is the preferential movement direction of the wheelchair, the descending direction of the potential field gradient is selected as the preferential movement direction of the wheelchair, and a plurality of obstacle avoidance paths can be found. The obstacle avoidance path refers to a path that the wheelchair can successfully reach. Calculating the sum SU of potential field values of all track points on the obstacle avoidance path, selecting the path with the smallest SU as an optimal path, optimizing the position of the point on the optimal path, adjusting the distance between the points or controlling the curvature change of the optimal path by using a track smoothing algorithm, and realizing the path smoothing and optimization to obtain a safe path.
The track points are continuously selected by selecting the descending direction of the potential field gradient according to the potential field value, so that the generated safety path is safer and more effective.
In summary, this embodiment provides a wheelchair control method, which integrates various sensors to realize intelligent perception and recognition of a wheelchair to the surrounding environment, and when the wheelchair is located and obstacle avoidance is needed, performs behavior prediction on a dynamic obstacle to calculate collision probability, and performs automatic obstacle avoidance to realize autonomous decision, autonomous navigation and intelligent obstacle avoidance of the wheelchair. The intelligent wheelchair can be better adapted to various complex environments, the obstacle avoidance and stability performances of the intelligent wheelchair are improved, the problem of frequent alarm is effectively reduced, and the use effect and the safety of the intelligent wheelchair are improved.
When the environmental risk level is low, although the risk of collision is low, the collision early warning function of the existing automatic navigation system may still generate early warning signals for several times. In order to avoid meaningless collision early warning and reduce the number of times of early warning, the invention closes the collision early warning function in the process of automatically avoiding the obstacle. But in order to enhance the safety, two time windows are set in a preset step length, and the collision probability of the navigation path is recalculated between the two time windows to prevent collision.
In one example, when the environmental risk level is low risk, the wheelchair is controlled to move along the safety path for a preset time step to automatically avoid the obstacle, as shown in fig. 3, and the specific steps include:
step S650: calculating a first time window according to the collision probability, and controlling the wheelchair to move along a safety path in the first time window;
first, a first time window Length1 (namely a first alarm-free time) is calculated according to the environmental risk levelWindow), the calculation formula is. Wherein P is the collision probability of the navigation path, and the range is 0-100%; a is a coefficient for correcting the influence degree of the quadratic term of P on the time window length, and the larger the value of a is, the more sensitive the window length is to the quadratic term change of P; b is a coefficient, and when a=0, the time window length is selected only by +.>Is determined by the linear term and constant c of (2); c is a constant term, the base part of the time window length. c mainly plays the following roles: when p=u=0, c determines the reference response time length of the wheelchair in the safest state, and the larger the value of c is, the more the wheelchair tends to automatically navigate for a long time in a safe environment, so that the better riding comfort can be provided; too small a value of c can frequently switch navigation modes, affecting experience. When the environment becomes complex (i.e. the P value becomes larger), c affects the magnitude of the variation of the first time window Length1, and the larger c is, the larger the time window Length is, the faster the response speed is. At this time, the magnitude of the c value needs to be determined by balancing the safety and comfort. Excessive c value can cause the first time window Length1 to oscillate or switch frequently, and the riding experience is poor. When the user intervenes actively (i.e. U gets larger), c also affects the measure of the increase of the first time window Length 1. The larger c is, the more the first time window Length1 is increased, and the active control request of the user is more easily met. However, too large c can also cause too large a fluctuation in the first time window Length1, and the system response is not smooth enough. When P and U change simultaneously and the effect is opposite, the larger the c value, the larger the influence of P on the change of the first time window Length 1. The environmental safety is more emphasized, and the response speed is faster. In summary, c is used as a constant basis of the first time window Length1, and the value of c directly influences the reference response time Length of the wheelchair control system in the safety environment, the response sensitivity to environment changes and the satisfaction degree of the active intervention request. Too large a value of c can lead to excessively sensitive or frequent switching of the control system response, influence the riding comfort, and too small a value of c can react slowly to environmental changes and is poor in safety.
U is a user habit parameter, the value is between 0 and 1, the preference of the user for selecting a time window is indicated, the higher the U value is, the longer the user hopes the time window, the longer the automatic experience time is, and the U value can be obtained according to the statistical result of long-term use data of the user. d is a user parameter correction coefficient, and represents the dependence degree of the first time window Length1 on the use habit of the user, and the larger the d value is, the larger the dependence of the first time window Length1 on the use habit of the user is. And the first time window Length1 is in seconds, generally 3-8 seconds, and if the value obtained according to the calculation formula exceeds the range, the boundary value is directly taken, so that the basic safety and the use experience are ensured.
Step S660: when the first time window is over, the collision probability of each dynamic obstacle and the navigation path is recalculated, and the environmental risk level is updated according to all the collision probabilities;
step S670: and when the updated environment risk level is low risk, controlling the wheelchair to move along the safety path within a preset second time window so as to prolong the automatic navigation time.
After the first time window is finished, returning to the step S500 to recalculate the collision probability, if the collision probability P of the navigation path calculated according to all the collision probabilities is still less than 0.3, indicating that the updated environmental risk level is still low risk, and controlling the wheelchair to move along the safety path in a preset second time window Length2 (i.e. a delay time window). The second time window Length2 is set for prolonging the automatic navigation time under the condition of lower and more stable environmental risks and improving the use efficiency. In general, the second time window Length2 > the first time window Length1, and the second time window Length2 may be set in the 9-20 second interval.
By the above, when the obstacle avoidance is performed automatically, the safety of the wheelchair can be ensured by closing the collision early warning function and setting the two-stage time window, and the user experience is better when the wheelchair moves.
When the automatic obstacle avoidance and automatic control are performed in the prolonged time window, the user may actively choose to exit the automatic control mode and enter the manual control mode due to the improvement of safety vigilance. The improvement of safety vigilance of the user indicates that the environmental risk is large, the automatic obstacle avoidance effect is poor or the subjective judgment of the user can not ensure the safety when the automatic obstacle avoidance is carried out, so that the user needs to switch to manual control. Because the effect of manual control is difficult to accurately judge in real time, a long time of observation is required to confirm the environmental change and the control effect, and the closing collision early warning is used for observing the automatic obstacle avoidance effect so as to determine when manual intervention is required. However, in the manual control mode, the user actively controls the device, so that the closing collision early warning loses the setting meaning, the effect is difficult to judge, and the wrong safety feeling can be provided in a certain time, so that the use vigilance is influenced. If the collision early warning is continuously closed, the safety of use is not guaranteed. Therefore, the collision early warning function needs to be started in the manual control mode.
After the manual control is finished, the collision probability P of the navigation path is calculated in real time, when the P is more than or equal to 0.3 and less than or equal to 0.5, the switching operation is prompted, the countdown is displayed, the user does not operate to directly enter the automatic navigation mode, the automatic mode can be restored only by manual confirmation after each active navigation is finished, and the interaction efficiency is improved; if P is more than or equal to 0.5, maintaining the manual control state.
Specifically, the implementation process of manual control is as follows: and simultaneously starting a brain wave acquisition device, an eye movement tracking device, a voice recognition and other multi-mode input modes to receive and recognize a control command of a user. Monitoring each input signal in real time, judging whether the input signals match with a preset operation intention template, if the eye movement change of a user shows that the intention of left turn is shown, the electroencephalogram signal shows the intention of deceleration, the voice command is 10 meters forward, and the like; and comprehensively judging control commands generated by the input of each mode, and determining the safest and effective control command in the current environment. If a command generated by one of the input modes causes a potential risk, the input mode is turned off. And outputting the generated control instruction to the control object to implement corresponding movement control. If the input of different modes cannot be effectively identified and matched, an obvious alarm prompt is sent out, and part of input modes are closed if necessary.
In one embodiment, when the wheelchair is controlled to move along the safety path within a preset second time window and an instruction of selecting the manual control mode by the user is received, as shown in fig. 4, the specific steps include:
step a671: collecting a first electroencephalogram signal of a user in real time;
step A672: when the user is judged to be in a vigilance state according to the first electroencephalogram signal and the pre-established electroencephalogram signal baseline, after a preset time interval, the second electroencephalogram signal of the user is acquired in real time, and if the user is judged to be still in the vigilance state according to the second electroencephalogram signal and the pre-established electroencephalogram signal baseline, the user responds to the instruction for confirming to enter a manual control mode, and the manual control mode is adopted to control the wheelchair to move and start a collision early warning function.
In this embodiment, normal electroencephalogram signals (i.e. electroencephalogram signals in a relaxed state relative to a vigilance state) of the user are collected, signal analysis is performed, and an electroencephalogram signal baseline of the user is established in advance. When the user selects to exit the automatic navigation mode through the control panel or the voice command, the first electroencephalogram signal of the user at the time is immediately acquired and analyzed. If it is detected that the beta electroencephalogram signal (neural excitation related) is significantly higher than the beta baseline in the electroencephalogram signal baselines, for example, more than 1.5 times in the normal state, which indicates that the vigilance and the attention level of the user are significantly improved, the user is determined to be in the vigilance state at this time. For further confirmation, the second electroencephalogram signal of the user is continuously acquired and analyzed, for example, after 3 seconds from the automatic mode exit, if the β electroencephalogram signal of the user is still significantly higher than the β baseline in the electroencephalogram signal baselines, the purpose that the user exits the automatic mode at this time is more likely to be subjective vigilance can be confirmed. The display prompt message "detect your vigilance state increase, switch to manual control mode. If the user chooses not to confirm, the system temporarily maintains the current mode, and continues to monitor the change of the electroencephalogram signal of the user to update the judgment. In the manual control mode, the system continues to detect and analyze the brain electrical signals of the user in real time. If the beta EEG signal is detected to be reduced to the normal level, judging that the subjective vigilance state of the user is reduced, and inquiring whether the automatic navigation mode needs to be switched or not according to prompt information. And if the user is not confirmed to switch to the automatic navigation mode by the two prompts, maintaining the manual control mode.
By comparing the electroencephalogram signal with the electroencephalogram signal baseline, whether the user selects the manual control mode due to the improvement of vigilance can be accurately judged.
In one embodiment, when the environmental risk level is low risk and an instruction of selecting the manual control mode is received when the wheelchair is controlled to move along the safety path within a preset second time window, as shown in fig. 5, the specific steps include:
step B671: counting the use time of an automatic navigation mode in a preset time period;
step B672: when the using time is greater than a preset threshold value, responding to a command of a user for confirming to enter a manual control mode, and controlling the wheelchair to move and start a collision early warning function in a preset third time window by adopting the manual control mode.
When the environmental risk level is low and the safety vigilance of the user is not improved, the reason that the user exits the automatic control mode is likely to be that the user hopes to manually control the experience, the user is given a certain time to adapt to the active control, and the experience requirement is met. Because of the safety period still in the second time window Lenth2, the manual control mode can set a third time window Length3, and the third time window Length3 is smaller than the second time window Length2 to ensure safety. Within the third time Length3 window, if a significant environmental change is detected, the third time window Length3 ends prematurely.
Specifically, the use time of the user in the automatic navigation mode within a preset period of time (e.g., the last 1 hour) is detected. If the fact that the service time of the user in the automatic navigation mode recently is longer than a preset threshold (for example, more than or equal to 30 minutes) is detected, the user is more customary to the automatic navigation mode; if the current environmental risk is low and the user is relatively adaptive to the automatic navigation mode, when the user is detected to select to exit the automatic navigation mode through the control panel or the voice command, the possible purpose of determining the user to exit the automatic navigation mode is more likely to be subjective manual control experience. At this time, prompt information is presented to the user: "detect you opt out of the automatic navigation mode, is you hope to switch to the manual handling experience mode: yes, manual control mode is started, no, EXIT. If the user selects ' yes ', the manual control mode ' is started, the manual control mode is entered, and meanwhile, a multi-mode control mode (such as brain-computer control, eye movement tracking and voice recognition) is started to receive a control instruction of the user. And controlling the wheelchair to move according to the control instruction of the user. If the user selects "no, EXIT," indicating that the purpose of exiting the automatic navigation mode this time is not for a manual manipulation experience. And exiting the judging process, and maintaining the current setting mode.
By the above, whether the reason for the user to exit the automatic control mode is that the user wants to control the experience manually or not is judged, so that the wheelchair control experience is better.
In one embodiment, in the process of automatically avoiding the obstacle, as shown in fig. 6, the method further comprises:
step S700: comparing the collision probability of the current preset time step with the collision probability of the previous preset time step to obtain the environmental risk change proportion;
in the automatic navigation mode, in each preset time step of the wheelchair automatic movement, according to the image continuously monitored by the environment sensor, the collision probability of the current preset time step navigation path is calculated, and according to the change proportion of the collision probability of the current preset time step navigation path compared with the collision probability of the previous time step navigation path, the environment risk change proportion is obtained. The collision probability of the navigation path with the previous preset time step is subtracted from the collision probability of the navigation path with the current preset time, and the obtained result is divided by the collision probability of the navigation path with the previous preset time step to obtain the environmental risk change proportion.
Step S800: and if the continuous environmental risk change proportion of the preset number exceeds the preset threshold, reducing the duration of the first time window and the second time window, and sending out an alarm prompt.
After continuously obtaining a plurality of environmental risk change ratios, if the continuous preset number of environmental risk change ratios all exceed a preset threshold, for example: when the continuous 3 environmental risk change ratios exceed 3%, the environmental risk is judged to be rapidly increased, at the moment, the duration of the automatic navigation time window (a first time window and a second time window) is shortened, an alarm prompt is sent in advance, the response delay caused by the environmental risk change is reduced to the maximum extent, and the use safety is ensured. If the environmental risk increasing trend is controlled in the subsequent new time window, the environmental risk changing proportion is stabilized in a safer range, and the duration of the time window of the automatic navigation is restored to the normal length. The wheelchair continues to travel along the safe path, and the environment sensor continues to monitor and position the environment until it eventually reaches the destination location and stops, waiting for the next instruction or new path planning.
By calculating the environmental risk change proportion according to the change of the collision probability of the navigation path, the environmental risk can be reflected from another angle, so that the wheelchair control is safer and more reliable.
Exemplary System
Based on the embodiment, the invention further provides an intelligent wheelchair. As shown in fig. 7, the intelligent wheelchair includes a processor, a memory, and a display screen connected through a system bus. Wherein the processor of the intelligent wheelchair is configured to provide computing and control capabilities. The memory of the intelligent wheelchair comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a smart wheelchair control program. The internal memory provides an environment for the operation of the operating system and the intelligent wheelchair control program in the non-volatile storage medium. The intelligent wheelchair control program, when executed by the processor, implements the steps of any one of the intelligent wheelchair control methods described above. The display screen of the intelligent terminal can be a liquid crystal display screen or an electronic ink display screen.
The embodiment of the invention also provides a computer readable storage medium, wherein the computer readable storage medium is stored with an intelligent wheelchair control program, and the intelligent wheelchair control program realizes the steps of any one of the intelligent wheelchair control methods provided by the embodiment of the invention when being executed by a processor.
It should be understood that the sequence number of each step in the above embodiment does not mean the sequence of execution, and the execution sequence of each process should be determined by its function and internal logic, and should not be construed as limiting the implementation process of the embodiment of the present invention.
The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art will understand that; the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions are not intended to depart from the spirit and scope of the various embodiments of the invention, which are also within the spirit and scope of the invention.
Claims (7)
1. The intelligent wheelchair control method is characterized by comprising the following steps:
Planning a navigation path based on the selected destination, and controlling the wheelchair to move along the navigation path by adopting an automatic navigation mode;
in the automatic navigation process, detecting an obstacle in an environment space where the wheelchair is positioned to obtain obstacle information;
obtaining the risk type of the position of the wheelchair according to the positions of all the obstacles in the obstacle information and the navigation path;
when the risk type is obstacle avoidance and obstacle avoidance is possible:
performing behavior prediction on the dynamic obstacle in the obstacle to obtain the position distribution information of the dynamic obstacle in a preset time step;
calculating collision probability of each dynamic obstacle and the navigation path according to the position distribution information, and obtaining an environmental risk level according to all the collision probabilities;
when the environmental risk level is high risk, controlling the wheelchair to move by adopting a manual control mode, otherwise, performing obstacle avoidance planning according to the obstacle information and the position distribution information to obtain a safety path, and controlling the wheelchair to move along the safety path within a preset time step so as to automatically avoid an obstacle;
and when the environmental risk level is low risk, closing a collision early warning function, and controlling the wheelchair to move along the safety path within a preset time step so as to automatically avoid the obstacle, wherein the method comprises the following steps of:
Calculating a first time window according to the collision probability, and controlling the wheelchair to move along the safety path in the first time window;
when the first time window is over, the collision probability of each dynamic obstacle and the navigation path is recalculated, and the environmental risk level is updated according to all the collision probabilities;
when the updated environment risk level is low risk, controlling the wheelchair to move along the safety path within a preset second time window so as to prolong the automatic navigation time;
when the wheelchair is controlled to move along the safety path in a preset second time window, and the manual control mode selection instruction is received by a user, the wheelchair further comprises:
counting the use time of an automatic navigation mode in a preset time period, and when the use time is larger than a preset threshold value, responding to a command of a user for confirming to enter a manual control mode, and controlling the wheelchair to move and start a collision early warning function by adopting the manual control mode in a preset third time window;
and acquiring a first electroencephalogram signal of the user in real time, when the user is judged to be in a vigilance state according to the first electroencephalogram signal and a pre-established electroencephalogram signal baseline, acquiring a second electroencephalogram signal of the user in real time after a preset time interval, and if the user is judged to be still in the vigilance state according to the second electroencephalogram signal and the pre-established electroencephalogram signal baseline, responding to an instruction of the user for confirming to enter a manual control mode, controlling the wheelchair to move and starting a collision early warning function in a manual control mode.
2. The intelligent wheelchair control method of claim 1, wherein the obtaining the risk type of the wheelchair location based on the location of all obstacles in the obstacle information and the navigation path comprises:
calculating a first distance between the position of each obstacle and the center of the wheelchair and a second distance between each obstacle and the navigation path;
when any one of the first distances is smaller than a first distance threshold, setting the risk type as an unavoidable obstacle;
otherwise, when all the second distances are larger than a second distance threshold, setting the risk type as not needing obstacle avoidance; and when any one of the second distances is smaller than or equal to a second distance threshold value, setting the risk type to be obstacle avoidance and obstacle avoidance.
3. The intelligent wheelchair control method of claim 1 wherein said calculating a collision probability of each of said dynamic obstacles with said navigation path based on said location distribution information comprises:
performing wheelchair movement simulation for preset times, and obtaining collision times of each dynamic obstacle and the navigation path according to the position distribution information;
and calculating the collision probability according to the category weight of each dynamic obstacle, the collision times, the preset times and the probability value in the position distribution information.
4. The intelligent wheelchair control method of claim 1, wherein the performing obstacle avoidance planning according to the obstacle information and the position distribution information to obtain a safety path comprises:
calculating potential field values of all positions of the navigation path region;
taking the direction of the decline of the potential field value as the moving direction of the wheelchair to obtain a plurality of obstacle avoidance paths;
accumulating potential field values of all track points on each obstacle avoidance path to obtain a total potential field value of each obstacle avoidance path;
and setting the obstacle avoidance path with the minimum total potential field value as the safety path.
5. The intelligent wheelchair control method of claim 1 further comprising, during the automatic obstacle avoidance process:
comparing the collision probability of the current preset time step with the collision probability of the previous preset time step to obtain an environmental risk change proportion;
and if the continuous preset number of environmental risk change ratios exceed a preset threshold, reducing the duration of the first time window and the second time window, and sending an alarm prompt.
6. A smart wheelchair comprising a memory, a processor and a smart wheelchair control program stored on the memory and operable on the processor, the smart wheelchair control program when executed by the processor implementing the steps of the smart wheelchair control method of any of claims 1-5.
7. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a smart wheelchair control program, which when executed by a processor, implements the steps of the smart wheelchair control method of any of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311141989.3A CN116880514B (en) | 2023-09-06 | 2023-09-06 | Intelligent wheelchair control method, intelligent wheelchair and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311141989.3A CN116880514B (en) | 2023-09-06 | 2023-09-06 | Intelligent wheelchair control method, intelligent wheelchair and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116880514A CN116880514A (en) | 2023-10-13 |
| CN116880514B true CN116880514B (en) | 2023-11-21 |
Family
ID=88255393
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311141989.3A Active CN116880514B (en) | 2023-09-06 | 2023-09-06 | Intelligent wheelchair control method, intelligent wheelchair and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116880514B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117277513B (en) * | 2023-11-15 | 2024-02-27 | 深圳安培时代数字能源科技有限公司 | Energy storage device control method, controller and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20080114127A (en) * | 2007-06-27 | 2008-12-31 | 한국과학기술원 | Wireless controllable modularized intelligent powered wheelchair |
| CN108478348A (en) * | 2018-05-29 | 2018-09-04 | 华南理工大学 | A kind of intelligent wheelchair and control method of interior independent navigation Internet of Things |
| CN112577491A (en) * | 2020-12-14 | 2021-03-30 | 上海应用技术大学 | Robot path planning method based on improved artificial potential field method |
| CN114647248A (en) * | 2022-05-23 | 2022-06-21 | 风林科技(深圳)有限公司 | Robot obstacle avoidance method, electronic device and readable storage medium |
| CN116570439A (en) * | 2023-07-13 | 2023-08-11 | 小舟科技有限公司 | Intelligent wheelchair control mode switching method and device, equipment and storage medium |
| CN116617011A (en) * | 2023-07-21 | 2023-08-22 | 小舟科技有限公司 | Wheelchair control method, device, terminal and medium based on physiological signals |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11730646B2 (en) * | 2020-09-25 | 2023-08-22 | Rajeev Ramanath | System and method to combine input from multiple sensors on an autonomous wheelchair |
-
2023
- 2023-09-06 CN CN202311141989.3A patent/CN116880514B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20080114127A (en) * | 2007-06-27 | 2008-12-31 | 한국과학기술원 | Wireless controllable modularized intelligent powered wheelchair |
| CN108478348A (en) * | 2018-05-29 | 2018-09-04 | 华南理工大学 | A kind of intelligent wheelchair and control method of interior independent navigation Internet of Things |
| CN112577491A (en) * | 2020-12-14 | 2021-03-30 | 上海应用技术大学 | Robot path planning method based on improved artificial potential field method |
| CN114647248A (en) * | 2022-05-23 | 2022-06-21 | 风林科技(深圳)有限公司 | Robot obstacle avoidance method, electronic device and readable storage medium |
| CN116570439A (en) * | 2023-07-13 | 2023-08-11 | 小舟科技有限公司 | Intelligent wheelchair control mode switching method and device, equipment and storage medium |
| CN116617011A (en) * | 2023-07-21 | 2023-08-22 | 小舟科技有限公司 | Wheelchair control method, device, terminal and medium based on physiological signals |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116880514A (en) | 2023-10-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7173105B2 (en) | AUTOMATIC OPERATION CONTROL DEVICE, AUTOMATIC OPERATION CONTROL METHOD, PROGRAM, AND AUTOMATIC OPERATION CONTROL SYSTEM | |
| US11772669B2 (en) | Processing a request signal regarding operation of an autonomous vehicle | |
| US11377108B2 (en) | Processing a request signal regarding operation of an autonomous vehicle | |
| EP2495713B1 (en) | Driving support device | |
| US11198436B2 (en) | Processing a request signal regarding operation of an autonomous vehicle | |
| KR102669020B1 (en) | Information processing devices, mobile devices, and methods, and programs | |
| KR102177826B1 (en) | Intervention in the operation of vehicles with autonomous driving capabilities | |
| US10850734B2 (en) | Processing a request signal regarding operation of an autonomous vehicle | |
| WO2020213280A1 (en) | Information processing device, mobile device, method, and program | |
| US20180284772A1 (en) | Processing a request signal regarding operation of an autonomous vehicle | |
| CN116880514B (en) | Intelligent wheelchair control method, intelligent wheelchair and storage medium | |
| CN110740916B (en) | Processing request signals related to operation of autonomous vehicles | |
| KR102288306B1 (en) | Wheelchair operation system and method for performing autonomous driving using map information linked with gis including wheelchair movement information | |
| US20180281795A1 (en) | Processing a request signal regarding operation of an autonomous vehicle | |
| US11040720B2 (en) | Sleepiness level prediction device and sleepiness level prediction method | |
| JP7308442B2 (en) | Information processing method, information processing device, and information processing system | |
| US20210256849A1 (en) | Process and system for local traffic approximation through analysis of cloud data | |
| CN116665155A (en) | Underground trackless man-vehicle getting-off early warning safety management system based on image processing technology | |
| CN113415278B (en) | System and process for following a recently on-road vehicle using surrounding vehicle motion flow | |
| EP3454269A1 (en) | Planning autonomous motion | |
| KR102178498B1 (en) | Wheelchair operation system and method for evaluating safety based on wheelchair's driving information and user's driving information | |
| JP2020192824A (en) | Driving behavior control method and driving behavior control device | |
| US20240161607A1 (en) | Event recording system, event recording device, and event recording method |
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 |