Depth learning-based disabled person travel path planning system and method

Abstract

The invention provides a disabled person travel path planning system and method based on deep learning, and the system comprises a data collection module which is used for collecting basic information, historical behavior data and position environment data of a disabled person; the behavior data analysis module is used for analyzing the basic information and historical behavior data of the disabled based on a deep learning algorithm to obtain features of the disabled; the environment data analysis module is used for processing the position environment data and obtaining environment characteristics of the position where the disabled are located; the path planning model building module is used for building a path planning model on the basis of the travel requirements of the disabled and in combination with the characteristics of the disabled and the environment characteristics; and the path planning module is used for providing a path planning scheme for the disabled user based on the path planning model. According to the technical scheme, more personalized, intelligent and dynamically adaptive path planning service can be realized, and more convenient, safer and more comfortable travel experience is provided for the disabled.

Description

Technical Field

[0001] The present invention belongs to the technical field of path planning, and specifically relates to a travel path planning system and method for disabled people based on deep learning. Background Art

[0002] With the rapid development of artificial intelligence technology, deep learning, as an important branch of machine learning, has achieved remarkable results in the fields of image recognition, speech recognition, natural language processing, etc. in recent years. By building a deep neural network model, deep learning can automatically learn and extract useful features from a large amount of data, providing new ideas and methods for solving complex problems. In the field of travel route planning for people with disabilities, deep learning technology can be used to analyze the travel habits, preferences and environmental factors of people with disabilities, thereby providing more personalized route planning services.

[0003] Traditional route planning systems often ignore the special needs of people with disabilities, such as the need for barrier-free facilities, adaptability of traffic flow, etc. Although some systems take the needs of people with disabilities into consideration, they lack personalized analysis and dynamic adjustment capabilities, making it difficult to meet the diverse needs of people with disabilities. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a travel route planning system and method for the disabled based on deep learning. The system has dynamic adjustment capabilities and can automatically optimize the route planning scheme according to the real-time needs of the disabled and environmental changes.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A travel path planning system for disabled people based on deep learning, the system comprising:

[0007] Data collection module, used to collect basic information, historical behavior data and location environment data of disabled people;

[0008] A behavior data analysis module is used to analyze the basic information and historical behavior data of disabled people based on a deep learning algorithm to obtain the characteristics of disabled people; wherein the characteristics of disabled people include disability type, disability level and behavior characteristics;

[0009] An environmental data analysis module, used to process the location environmental data to obtain environmental characteristics of the location where the disabled person is located; wherein the environmental characteristics include traffic characteristics, map characteristics and meteorological characteristics;

[0010] A path planning model building module is used to establish a path planning model based on the travel needs of the disabled and in combination with the characteristics of the disabled and the environmental characteristics;

[0011] The path planning module is used to provide a path planning solution for disabled users based on the path planning model.

[0012] Preferably, the behavior data analysis module includes:

[0013] A behavior data processing unit is used to clean the basic information and historical behavior data of the disabled and obtain the cleaned data;

[0014] A model building unit, used to build a deep learning model based on a capsule network, and extract features from the cleaned data based on the deep learning model to obtain the features of the disabled person;

[0015] The capsule network includes a convolutional layer for extracting first-order disabled person features, a capsule layer for extracting second-order disabled person features, and an output layer for outputting final disabled person features; the capsule layer includes a primary capsule layer and a digital capsule layer.

[0016] Preferably, the model building unit comprises:

[0017] The squeeze-excitation building subunit is used to build a squeeze layer based on global average pooling and a stimulus layer based on a fully connected layer;

[0018] An extrusion-stimulation-insertion subunit, used for inserting the extrusion layer and the stimulation layer into the primary capsule layer to obtain an improved primary capsule layer;

[0019] A feature extraction subunit, used for inputting the first-order disabled person features extracted by the convolution layer into the improved primary capsule layer to obtain weighted second-order disabled person features;

[0020] A dynamic routing sub-unit, for dynamically allocating connection weights to the improved primary capsule layer and the digital capsule layer based on a dynamic routing mechanism, inputting the weighted second-order disabled person features into the digital capsule layer, and obtaining a disabled person feature vector;

[0021] The feature output subunit is used to output the final disabled person feature based on the disabled person feature vector.

[0022] Preferably, in the environmental data analysis module,

[0023] The traffic characteristics include road flatness and width, distribution of barrier-free facilities, public transportation conditions, traffic flow and congestion;

[0024] The map features include latitude and longitude and altitude, topography and slope, building distribution and green space coverage;

[0025] The meteorological characteristics include temperature and humidity, precipitation conditions, and wind speed and direction.

[0026] Preferably, the path planning model building module includes:

[0027] A population initialization unit, used to randomly generate a set of initial path planning solutions as an initial population of the genetic algorithm according to the travel needs of the disabled and the characteristics of the disabled;

[0028] A fitness evaluation unit, used to evaluate the fitness of each initial path planning scheme according to the environmental characteristics and the characteristics of the disabled person; wherein the evaluation indicators include path length, barrier-free facility coverage and travel time;

[0029] A selection unit is used to select an initial path planning scheme whose fitness meets a preset threshold as a parent generation according to the fitness evaluation result, so as to generate the next generation of offspring;

[0030] The crossover and mutation unit is used to perform crossover and mutation operations on the selected parent generation, and introduce a weight perturbation mechanism to dynamically adjust the weights of different environmental features according to the characteristics of the disabled person, and generate a new path planning scheme for the offspring generation;

[0031] The local search unit is used to perform local search for each new child path planning scheme generated using a simulated annealing algorithm to obtain the fitness value of the new child path planning scheme, and take the child path planning scheme whose fitness value meets the preset target fitness value as the optimal path planning scheme to complete the construction of the path planning model.

[0032] Preferably, the system also includes a reinforcement learning module for collecting travel feedback data of disabled people, constructing a reward function based on the feedback data, and optimizing the path planning model through the reward function.

[0033] The present invention also provides a travel path planning method for disabled people based on deep learning, and the system is applied, including:

[0034] Collect basic information, historical behavior data, and location and environmental data of disabled persons;

[0035] Based on the deep learning algorithm, the basic information and historical behavior data of the disabled are analyzed to obtain the characteristics of the disabled; wherein the characteristics of the disabled include the type of disability, the degree of disability and the behavioral characteristics;

[0036] Processing the location environment data to obtain the environment characteristics of the location where the disabled person is located; wherein the environment characteristics include traffic characteristics, map characteristics and meteorological characteristics;

[0037] Based on the travel needs of the disabled, combined with the characteristics of the disabled and the environmental characteristics, a path planning model is established;

[0038] Based on the path planning model, a path planning solution is provided for disabled users.

[0039] Preferably, the method for obtaining the characteristics of disabled persons comprises:

[0040] Clean the basic information and historical behavior data of disabled people to obtain cleaned data;

[0041] Constructing a deep learning model based on a capsule network, and performing feature extraction on the cleaned data based on the deep learning model to obtain the features of the disabled person;

[0042] The capsule network includes a convolutional layer for extracting first-order disabled person features, a capsule layer for extracting second-order disabled person features, and an output layer for outputting final disabled person features; the capsule layer includes a primary capsule layer and a digital capsule layer.

[0043] Compared with the prior art, the beneficial effects of the present invention are as follows: the data acquisition module provides a rich data foundation for the system, which is helpful for subsequent analysis and route planning. Through the deep learning algorithm of the behavioral data analysis module, the system can more accurately understand the travel needs and preferences of the disabled, and provide more accurate guidance for subsequent route planning. The environmental data analysis module provides comprehensive environmental information support for route planning, which helps to plan a safer and more convenient travel route. By constructing a special route planning model, the route planning model construction module enables the system to provide more personalized and intelligent travel services for the disabled. In summary, the technical solution of the present invention accurately analyzes the characteristics of the disabled through a deep learning algorithm, and realizes personalized route planning. The impact of environmental characteristics on route planning is comprehensively considered to improve the feasibility and safety of the route. The system has dynamic adjustment capabilities and can automatically optimize the route planning scheme according to the real-time needs of the disabled and environmental changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0045] Figure 1 This is a schematic diagram of the structure of a travel route planning system for the disabled based on deep learning in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] Embodiment 1

[0049] like Figure 1 As shown, a travel path planning system for disabled people based on deep learning, the system includes: a data acquisition module, a behavior data analysis module, an environmental data analysis module, a path planning model construction module and a path planning module;

[0050] The data collection module is used to collect basic information, historical behavior data, and location and environment data of disabled people; among them, the basic information of disabled people includes name, age, gender, medical history, and disability type (such as visual impairment, hearing impairment, physical disability, etc., which are key factors that must be considered when planning barrier-free travel routes.). Historical behavior data includes historical travel records (travel time, starting point, end point, etc.), preference settings (travel frequency, route selection, transportation mode, etc.). Location and environment data include real-time or predicted traffic flow, congestion, traffic accidents, road layout, intersections, sidewalks, detailed locations of barrier-free facilities (such as blind paths, ramps), rain, snow, strong winds, high temperatures, etc., and the location and availability of elevators, barrier-free toilets, wheelchair ramps, etc. By collecting a large amount of multi-source heterogeneous data, data support is provided for path planning.

[0051] The behavioral data analysis module is used to analyze the basic information and historical behavioral data of disabled people based on deep learning algorithms to obtain the characteristics of disabled people; among which, the characteristics of disabled people include disability type, disability level and behavioral characteristics.

[0052] In a further embodiment, the behavior data analysis module includes:

[0053] The behavior data processing unit is used to clean the basic information and historical behavior data of the disabled to obtain the cleaned data; specifically, in this embodiment, according to the pre-established data mapping rules and standardized templates, the cleaned historical behavior heterogeneous data is subjected to field mapping and format conversion, and the data from different sources are converted into a unified standardized format to form a structured standard data set. Data fusion technology is used to deduplicate and merge the standardized data, and multiple data records describing the same entity are integrated into a complete data record to improve the integrity and consistency of the data. Through data association analysis, the association relationship between different data tables and fields is mined, and the association map between data is constructed to provide support for subsequent data analysis and application.

[0054] A model building unit, used to build a deep learning model based on a capsule network, and to extract features from the cleaned data based on the deep learning model to obtain features of disabled people;

[0055] Among them, the capsule network is a set of neurons, including a convolutional layer for extracting first-order disabled people's features, a capsule layer for extracting second-order disabled people's features, and an output layer for outputting the final disabled people's features; the capsule layer includes a primary capsule layer and a digital capsule layer.

[0056] A further embodiment is that the model building unit comprises:

[0057] The squeeze-excitation building subunit is used to build a squeeze layer based on global average pooling and a stimulus layer based on a fully connected layer;

[0058] An extrusion-excitation-insertion subunit is used to insert an extrusion layer and an excitation layer into the primary capsule layer to obtain an improved primary capsule layer;

[0059] A feature extraction subunit, used for inputting the first-order disabled person features extracted by the convolution layer into the improved primary capsule layer to obtain weighted second-order disabled person features;

[0060] The dynamic routing subunit is used to dynamically allocate connection weights for the improved primary capsule layer and the digital capsule layer based on the dynamic routing mechanism, input the weighted second-order disabled person features into the digital capsule layer, and obtain the disabled person feature vector; in this embodiment, the connection weights are calculated and the output of the digital capsule is updated according to the input second-order features and the current state of the digital capsule. Specifically, the first-order disabled person features are input into the primary capsule layer and rearranged to obtain the reconstructed first-order disabled person features. The reconstructed first-order disabled person features are multiplied by their corresponding weights to obtain a prediction vector matrix, and the prediction vector is weighted to obtain the second-order disabled person features. These second-order features not only contain the information of local features, but also emphasize the parts related to the disabled person features through weighted processing. Among them, the coupling coefficient between the primary capsule layer and the digital capsule layer is integrated as 1, which determines the performance of the capsule network routing algorithm. After multiple iterations, the digital capsule layer will output a feature vector, which represents the overall features of the disabled person. The length and direction of the feature vector represent the probability of existence and specific attributes of the disabled person features, respectively. In particular, the digital capsule layer obtained after multiple routings is divided into two branches, one of which is used to calculate the length of the feature classifier and the other branch is used to reconstruct the original data.

[0061] The feature output subunit is used to output the final disabled person features based on the disabled person feature vector.

[0062] An environmental data analysis module is used to process the location environmental data to obtain the environmental characteristics of the location where the disabled person is located; wherein the environmental characteristics include traffic characteristics, map characteristics and meteorological characteristics; a further implementation method is that in the environmental data analysis module, the traffic characteristics include road flatness and width, distribution of barrier-free facilities, public transportation conditions, traffic flow and congestion; map characteristics include latitude and longitude and altitude, terrain and slope, building distribution and green space coverage;

[0063] Meteorological characteristics include temperature and humidity, precipitation conditions, and wind speed and direction.

[0064] The path planning model building module is used to establish a path planning model based on the travel needs of the disabled and in combination with the characteristics of the disabled and the environment. A further implementation method is that the path planning model building module includes:

[0065] A population initialization unit is used to randomly generate a set of initial path planning solutions as the initial population of the genetic algorithm according to the travel needs and characteristics of the disabled;

[0066] The fitness evaluation unit is used to evaluate the fitness of each initial path planning scheme according to the environmental characteristics and the characteristics of the disabled; wherein the evaluation indicators include the path length, barrier-free facility coverage and travel time; in this embodiment, the traffic environment factors (road congestion, accident conditions, etc.) are introduced into the evaluation indicators to perform weight perturbation on the evaluation indicators, and the temperature annealing factor in the simulated annealing algorithm is incorporated at the same time, and the fitness function is adaptively changed by controlling the temperature annealing factor. In particular, the weight perturbation of the evaluation indicators is to construct a user profile according to the disability characteristics of the current disabled user, and based on the user profile, different weight allocation schemes can be preset for different types of disabled users.

[0067] A selection unit is used to select an initial path planning scheme whose fitness meets a preset threshold as a parent generation according to the fitness evaluation result, so as to generate the next generation of offspring;

[0068] The crossover and mutation unit is used to perform crossover and mutation operations on the selected parent generation, and introduce a weight perturbation mechanism to dynamically adjust the weights of different environmental characteristics according to the characteristics of the disabled, and generate a new offspring path planning scheme; specifically, based on the constant factor, the heuristic mutation factor, and the preset upper and lower limits of the adaptive mutation probability, the adaptive mutation probability is calculated. In this embodiment, the temperature annealing factor is introduced into the calculation formula of the adaptive mutation probability. When the fitness value of the mutant individual is greater than the average fitness value and the temperature annealing factor is greater than the preset critical value, the adaptive mutation probability is calculated. Whether the parent generation mutates is controlled based on the adaptive mutation probability.

[0069] The local search unit is used to perform local search for each new child path planning scheme generated using a simulated annealing algorithm to obtain the fitness value of the new child path planning scheme, and take the child path planning scheme whose fitness value meets the preset target fitness value as the optimal path planning scheme to complete the construction of the path planning model.

[0070] The path planning module is used to provide path planning solutions for disabled users based on the path planning model.

[0071] A further implementation method is that the system also includes a reinforcement learning module for collecting travel feedback data of disabled people, and constructing a reward function based on the feedback data, and optimizing the path planning model through the reward function.

[0072] In this embodiment, the path planning model is continuously optimized based on user feedback:

[0073] First, define the state space and action space, where the state space includes the current location, destination, traffic conditions, environmental characteristics, etc.; the action space includes possible path choices.

[0074] Then, a reward function is designed based on user feedback (satisfaction score, frequency of choosing a certain path). For example, if the user prefers to avoid crowded areas, a positive reward is given when the path recommended by the model avoids crowded areas. The reinforcement learning module continuously optimizes the path planning model through policy iteration. In each iteration, the module selects a path based on the current policy and collects feedback data to update the reward function. Based on the feedback from the reward function, the module adjusts the parameters of the path planning model to maximize the cumulative reward. This process is repeated until the model converges to the optimal solution or reaches a predetermined number of iterations.

[0075] Because the travel trajectory data of disabled users contains information such as location coordinates, timestamps, and travel purposes. Taking wheelchair users as an example, the starting and ending points, route selection, time consumption, and other data of each trip are recorded through mobile phone applications. Users can rate and evaluate each travel experience, such as whether the route is convenient and whether the barrier-free facilities are complete. These data constitute the basic data set for model optimization. In the process of model optimization, the reward function optimization method is used. For example, if a user reports that the sidewalk of a certain path is too narrow and not suitable for wheelchairs, a negative reward function is generated after the user's feedback. The system will reduce the weight of the section in the path planning according to the negative reward function. With the continuous accumulation of feedback data, the model is gradually optimized, and the recommendation effect continues to improve. By analyzing the user's historical travel data, their travel preferences can be discovered. For example, a visually impaired user is accustomed to choosing a section with a voice prompt device, and the system will use this preference as an important reference for path planning. If a user often goes to a certain place at 8 o'clock in the morning, the system will record this rule and actively recommend a suitable route during the corresponding period. User satisfaction evaluation directly affects the path recommendation strategy. The satisfaction threshold is set to 3.5 points on a five-point scale. If the average score of a path is lower than the threshold, the system will reduce its recommendation frequency accordingly. For example, if a path is temporarily diverted due to construction and the user score is low, the system will adjust the recommendation strategy in time. The collaborative filtering algorithm finds similar user groups by analyzing user characteristics. For example, users who use wheelchairs and often move in the same area have similar travel path choices. The system will select the route suitable for the current user from the high-scoring paths of similar users for recommendation. When considering multiple factors, the system assigns different weights to different factors. The weight of the completeness of barrier-free facilities is 0.4, the weight of the density of people flow is 0.3, and the weight of the slope of the road section is 0.3. For example, although a commercial area has dense traffic, it has complete barrier-free facilities and a gentle slope, and it may still be recommended by the system. The system continuously tracks the use of the path, records whether the user drives along the recommended route, whether the route is changed midway, and whether the actual time consumption meets expectations. These data are used to evaluate the recommendation effect and serve as an important basis for model optimization. When it is found that the actual travel time of a recommended route generally exceeds the estimated value, the system will adjust the time estimation model accordingly.

[0076] Embodiment 2

[0077] The present invention also provides a travel path planning method for disabled persons based on deep learning, and an application system, comprising:

[0078] Collect basic information, historical behavior data, and location and environmental data of disabled persons;

[0079] Based on deep learning algorithms, the basic information and historical behavior data of disabled people are analyzed to obtain the characteristics of disabled people, including disability type, disability level and behavioral characteristics.

[0080] Processing the location environment data to obtain the environment characteristics of the location where the disabled person is located; wherein the environment characteristics include traffic characteristics, map characteristics and meteorological characteristics;

[0081] Based on the travel needs of people with disabilities, combined with the characteristics of people with disabilities and environmental characteristics, a path planning model is established;

[0082] Based on the path planning model, path planning solutions are provided for disabled users.

[0083] A further implementation method is that the method for obtaining characteristics of a disabled person includes:

[0084] Clean the basic information and historical behavior data of disabled people to obtain cleaned data;

[0085] Construct a deep learning model based on capsule network, and extract features from the cleaned data based on the deep learning model to obtain the characteristics of disabled people;

[0086] Among them, the capsule network includes a convolution layer for extracting first-order disabled people's features, a capsule layer for extracting second-order disabled people's features, and an output layer for outputting final disabled people's features; the capsule layer includes a primary capsule layer and a digital capsule layer.

[0087] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A travel path planning system for disabled people based on deep learning, characterized in that: The system comprises: Data collection module, used to collect basic information, historical behavior data and location environment data of disabled people; A behavior data analysis module is used to analyze the basic information and historical behavior data of disabled people based on a deep learning algorithm to obtain the characteristics of disabled people; wherein the characteristics of disabled people include disability type, disability level and behavior characteristics; An environmental data analysis module, used to process the location environmental data to obtain environmental characteristics of the location where the disabled person is located; wherein the environmental characteristics include traffic characteristics, map characteristics and meteorological characteristics; A path planning model building module is used to establish a path planning model based on the travel needs of the disabled and in combination with the characteristics of the disabled and the environmental characteristics; The path planning module is used to provide a path planning solution for disabled users based on the path planning model.

2. The travel path planning system for disabled people based on deep learning according to claim 1 is characterized in that: The behavioral data analysis module includes: A behavior data processing unit is used to clean the basic information and historical behavior data of the disabled and obtain the cleaned data; A model building unit, used to build a deep learning model based on a capsule network, and extract features from the cleaned data based on the deep learning model to obtain the features of the disabled person; The capsule network includes a convolutional layer for extracting first-order disabled person features, a capsule layer for extracting second-order disabled person features, and an output layer for outputting final disabled person features; the capsule layer includes a primary capsule layer and a digital capsule layer.

3. The deep learning-based travel path planning system for disabled people according to claim 2 is characterized in that: The model building unit comprises: The squeeze-excitation building subunit is used to build a squeeze layer based on global average pooling and a stimulus layer based on a fully connected layer; An extrusion-stimulation-insertion subunit, used for inserting the extrusion layer and the stimulation layer into the primary capsule layer to obtain an improved primary capsule layer; A feature extraction subunit, used for inputting the first-order disabled person features extracted by the convolution layer into the improved primary capsule layer to obtain weighted second-order disabled person features; A dynamic routing sub-unit, for dynamically allocating connection weights to the improved primary capsule layer and the digital capsule layer based on a dynamic routing mechanism, inputting the weighted second-order disabled person features into the digital capsule layer, and obtaining a disabled person feature vector; The feature output subunit is used to output the final disabled person feature based on the disabled person feature vector.

4. The travel path planning system for disabled people based on deep learning according to claim 2 is characterized in that: In the environmental data analysis module, The traffic characteristics include road flatness and width, distribution of barrier-free facilities, public transportation conditions, traffic flow and congestion; The map features include latitude and longitude and altitude, topography and slope, building distribution and green space coverage; The meteorological characteristics include temperature and humidity, precipitation conditions, and wind speed and direction.

5. The travel path planning system for disabled people based on deep learning according to claim 2 is characterized in that: The path planning model building module includes: A population initialization unit, used to randomly generate a set of initial path planning solutions as an initial population of the genetic algorithm according to the travel needs of the disabled and the characteristics of the disabled; A fitness evaluation unit, used to evaluate the fitness of each initial path planning scheme according to the environmental characteristics and the characteristics of the disabled person; wherein the evaluation indicators include path length, barrier-free facility coverage and travel time; A selection unit is used to select an initial path planning scheme whose fitness meets a preset threshold as a parent generation according to the fitness evaluation result, so as to generate the next generation of offspring; The crossover and mutation unit is used to perform crossover and mutation operations on the selected parent generation, and introduce a weight perturbation mechanism to dynamically adjust the weights of different environmental features according to the characteristics of the disabled person, and generate a new path planning scheme for the offspring generation; The local search unit is used to perform local search for each new child path planning scheme generated using a simulated annealing algorithm to obtain the fitness value of the new child path planning scheme, and take the child path planning scheme whose fitness value meets the preset target fitness value as the optimal path planning scheme to complete the construction of the path planning model.

6. The travel path planning system for disabled people based on deep learning according to claim 2 is characterized in that: The system also includes a reinforcement learning module for collecting travel feedback data of disabled people, constructing a reward function based on the feedback data, and optimizing the path planning model through the reward function.

7. A method for planning travel routes for disabled persons based on deep learning, using the system described in any one of claims 1 to 6, characterized in that: include: Collect basic information, historical behavior data, and location and environmental data of disabled persons; Based on the deep learning algorithm, the basic information and historical behavior data of the disabled are analyzed to obtain the characteristics of the disabled; wherein the characteristics of the disabled include the type of disability, the degree of disability and the behavioral characteristics; Processing the location environment data to obtain the environment characteristics of the location where the disabled person is located; wherein the environment characteristics include traffic characteristics, map characteristics and meteorological characteristics; Based on the travel needs of the disabled, combined with the characteristics of the disabled and the environmental characteristics, a path planning model is established; Based on the path planning model, a path planning solution is provided for disabled users.

8. The method for planning travel paths for disabled persons based on deep learning according to claim 7 is characterized in that: Methods for obtaining disability characteristics include: Clean the basic information and historical behavior data of disabled people to obtain cleaned data; Constructing a deep learning model based on a capsule network, and performing feature extraction on the cleaned data based on the deep learning model to obtain the features of the disabled person; The capsule network includes a convolutional layer for extracting first-order disabled person features, a capsule layer for extracting second-order disabled person features, and an output layer for outputting final disabled person features; the capsule layer includes a primary capsule layer and a digital capsule layer.

Images