An obstacle avoidance learning method for tracked vehicles in coal mines

Abstract

The invention discloses an obstacle avoidance learning method for a tracked vehicle in an underground coal mine, comprising the following steps: the invention establishes a tracked vehicle obstacle avoidance database and a basic sub-database for obstacle avoidance of the vehicle, and when receiving data, compares and analyzes the data in the database, thereby screening and extracting obstacle avoidance information that meets the requirements for one-to-one matching, and then records the data in the tracked vehicle system, and simultaneously performs multiple simulation tests on the tracked vehicle to obtain driving data, path levels, and collision times and strengths of the tracked vehicle in the simulation scene, thereby not only reducing the data processing volume of the tracked vehicle, but also avoiding the long driving distance and heavy travel tasks of the tracked vehicle without path planning, which is beneficial to practical application and operation.

Description

Technical Field

[0001] The invention belongs to the technical field of tracked vehicle obstacle avoidance, and in particular relates to an obstacle avoidance learning method for a tracked vehicle in an underground coal mine. Background Art

[0002] In our daily life, coal mines are areas where humans mine coal resources in coal-rich mining areas. They are generally divided into underground coal mines and open-pit coal mines. When the coal seam is far from the surface, people generally choose to dig tunnels underground to mine coal. This is an underground coal mine. When the coal seam is very close to the surface, people generally choose to directly strip the surface soil layer to mine coal. This is an open-pit coal mine. Most of my country's coal mines are underground coal mines. The scope of coal mines includes a large area above and below the ground and related facilities. Coal mines are dug by humans when digging geological strata rich in coal. Reasonable space usually includes tunnels, shafts and mining faces, etc. Coal is the main solid fuel and a kind of combustible organic rock. It is formed by lush plants growing in a certain geological period. In a suitable geological environment, it gradually accumulates into thick layers and is buried under water or in mud. It is formed by natural coalification over a long geological period. A coal mine is a general term for shafts, chambers, equipment, ground buildings and structures that form an underground coal mine production system. A coal mine is a shaft dug by workers to develop coal mines.

[0003] However, existing coal mines often have small entrances, and workers are unable to survey the complex coal mine shafts. At this time, tracked vehicles are needed to survey the deep layers of the coal mines and provide detection data, and upload it to operators on the ground to analyze the data in the mine. If effective and real data inside the mine cannot be obtained, it is easy to make wrong judgments and predictions about the coal mine, which greatly reduces the safety of the workers and is not conducive to practical application and operation. Summary of the invention

[0004] The purpose of the present invention is to provide an obstacle avoidance learning method for a crawler vehicle in an underground coal mine in order to solve the above-mentioned problems, thereby solving the problems mentioned in the background technology.

[0005] In order to solve the above problems, the present invention provides a technical solution:

[0006] A method for learning obstacle avoidance for a crawler vehicle in a coal mine, comprising the following steps:

[0007] S1. Establish a tracked vehicle obstacle avoidance database for storing the tracked vehicle obstacle avoidance program, and establish a basic sub-database for the tracked vehicle obstacle avoidance based on the data information of other tracked vehicle databases for storing the obstacle avoidance programs of different tracked vehicles in other environments;

[0008] S2, the crawler vehicle sensor system identifies the real-time acquired environmental information and transmits it to the control center through the communication circuit. The control center analyzes the sensor data and outputs the environmental information in a standard format;

[0009] S3, then detect the moving obstacles in the environment during driving, which is completed by the vehicle-mounted environmental perception system, and then rate and predict the possibility of obstacles that may be encountered during driving, judge the collision relationship with obstacles, and make the tracked vehicle avoid obstacles safely through intelligent decision-making and path planning, which is determined by the tracked vehicle path decision system;

[0010] S4. Simultaneously, a simulation test is performed on the crawler vehicle to simulate the obstacle avoidance path and performance of the crawler vehicle in the coal mine. If the obstacle avoidance effect is not good, the database of the crawler vehicle will be program-changed or optimized, and the crawler vehicle will be simulated again until the obstacle avoidance rate of the crawler vehicle is 100%;

[0011] S5. Place the crawler vehicle in a real coal mine for simulation testing. After the test, optimize the program until the crawler vehicle successfully avoids danger, so that the crawler vehicle can adapt to the complex real environment of the coal mine.

[0012] Preferably, the specific operation steps of establishing the big data software information database in step S1 are:

[0013] S101, collecting a database of tracked vehicle obstacle avoidance, and collecting all tracked vehicle category information and all tracked vehicle information under various categories through a tracked vehicle data center;

[0014] S102, respectively establish all obstacle avoidance information databases, and count the obstacle avoidance information belonging to each category into the corresponding tracked vehicle classification sub-data layer to form a tracked vehicle information total database;

[0015] S103, recording the first letters of the obstacle avoidance sub-data layer, extracting the first letter of each first letter in the sub-database, and arranging them in alphabetical order, grouping the first letters of each obstacle avoidance database under the same first letter, forming an alphabetical index of the obstacle avoidance database information, which is used to modify and optimize the information in the database, and finally forming a standard obstacle avoidance database:

[0016] S111, extracting all obstacle avoidance information from each tracked vehicle database, and recording the first letters of the obstacle avoidance data of each tracked vehicle;

[0017] S112, arranging the first letters of the first words of various obstacle avoidance information in alphabetical order, summarizing the first letters of the information in the tracked vehicle data, and creating an information folder under the corresponding obstacle avoidance information to form a basic obstacle avoidance sub-database;

[0018] S113, and use the first English letter of the first word of the information in each obstacle avoidance database to sort them in order, and form an index in each obstacle avoidance sub-database.

[0019] Preferably, the tracked vehicle obstacle avoidance database in step S1 is interactively connected with each obstacle avoidance basic sub-database via a data hierarchy.

[0020] Preferably, the specific operation steps of the sensing system identifying the environmental information acquired in real time in step S2 are:

[0021] S21. The sensing system includes an ultrasonic sensor, an infrared sensor, an extended encoder, a visual sensor, a program modifier and a collision sensor, wherein the collision sensor and the extended encoder are both priority sensors used to handle obstacle avoidance situations; the infrared sensor, the ultrasonic sensor and the visual sensor are all low-priority sensors, which are obstacle avoidance means for detecting the distance and path planning and category of the current road section and performing corresponding processing, so as to realize the step of face-to-face and real-time obstacle avoidance operation for the current obstacle; and the program modifier is a necessary means for the tracked vehicle to be unable to avoid obstacles after detection, and the program that cannot avoid obstacles is modified and optimized to achieve a 100% success rate of obstacle avoidance;

[0022] S22. Transmitting data to the control center according to the communication circuit, and then the control center performs hierarchical analysis on the obstacle avoidance data, so as to obtain environmental information in a standard format, so that the control center can use it as a standard for subsequent execution.

[0023] Preferably, the specific operation steps of possibility rating and prediction in step S3 are:

[0024] S31. Under the vehicle-mounted system perception system, the matched scene model is compared with the information in the standardized obstacle avoidance database one by one, so as to rate and predict the obstacles encountered by the tracked vehicle during the current driving process. If the obstacle is huge, multiple routes need to be planned and the travel speed is slow, the rating is S. If there are many obstacles, the routes need to cross and multi-angle adjustment and multi-sensing are required in the middle, the rating is A. If the road is flat during driving and no other emergency measures are required to climb over the obstacles, the rating is B.

[0025] S32. At the same time, the sensing system inside the crawler vehicle will pre-plan and reorganize the path, making the best choice for the subsequent multiple obstacle sections.

[0026] Preferably, the specific operation steps of performing the simulation test on the crawler vehicle in step S4 are:

[0027] Combined with the existing obstacle avoidance data information, the tracked vehicle will be driven in a specific environment, and sensors and gravity sensors will be inserted into the obstacle simulators. The number and intensity of the tracked vehicle's collisions with obstacles during driving will be recorded, and the real-time data will be uploaded to the obstacle avoidance big data background. Computer algorithms will be used to compare the multiple collision data before and after, and the tracked vehicle program will be optimized and modified.

[0028] Preferably, the specific operation steps of performing a real simulation test on the crawler vehicle in step S5 are:

[0029] Based on the number of times the tracked vehicle avoids obstacles and the intensity of impact in the simulated environment, the tracked vehicle is placed in a real coal mine background for simulation testing. The sensing system will upload the real data to the cloud, and the sensing system will plan the travel path and apply it to this driving journey until the tracked vehicle avoids the obstacle and passes it completely.

[0030] The beneficial effects of the present invention are as follows: the present invention establishes a tracked vehicle obstacle avoidance database and a basic sub-database for obstacle avoidance of the vehicle, and when receiving data, compares and analyzes the data in the database, thereby screening and extracting obstacle avoidance information that meets the requirements for one-to-one matching, and then recording the data in the tracked vehicle system, and performing multiple simulation tests on the tracked vehicle to obtain the driving data, path level, and number and intensity of collisions of the tracked vehicle in the simulation scene, which not only reduces the data processing volume of the tracked vehicle, but also avoids the long driving distance and heavy travel tasks of the tracked vehicle without path planning, which is beneficial to practical application and operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] For ease of explanation, the present invention is described in detail with reference to the following specific implementations and the accompanying drawings.

[0032] Figure 1 The present invention is based on a flow chart of an obstacle avoidance learning method for a tracked vehicle in an underground coal mine. DETAILED DESCRIPTION

[0033] like Figure 1 As shown, this specific implementation adopts the following technical solutions:

[0034] Example:

[0035] A method for learning obstacle avoidance for a crawler vehicle in a coal mine, comprising the following steps:

[0036] S1. Establish a tracked vehicle obstacle avoidance database for storing tracked vehicle obstacle avoidance programs, and establish a basic sub-database for obstacle avoidance of this vehicle based on data information in other tracked vehicle databases, for storing obstacle avoidance programs of different tracked vehicles in other environments, so as to facilitate the collection of obstacle avoidance programs of other models;

[0037] S2. The crawler vehicle sensor system identifies the real-time acquired environmental information and transmits it to the control center through the communication circuit. The control center analyzes the sensor data and outputs the environmental information in a standard format to facilitate subsequent analysis based on this standard.

[0038] S3. Detect moving obstacles in the environment during driving, which is completed by the vehicle-mounted environmental perception system. Then, the possibility of obstacles that may be encountered during driving is rated and predicted, and the collision relationship with obstacles is judged. Through intelligent decision-making and path planning, the tracked vehicle can safely avoid obstacles. The tracked vehicle path decision system determines the collision relationship and plans the priority path.

[0039] S4. Simultaneously, a simulation test is performed on the crawler vehicle to simulate the obstacle avoidance path and performance of the crawler vehicle in the coal mine. If the obstacle avoidance effect is not good, the database of the crawler vehicle will be program-changed or optimized, and the crawler vehicle will be simulated again until the obstacle avoidance rate of the crawler vehicle is 100%, so as to facilitate data collection and optimization;

[0040] S5. The crawler vehicle is then placed in a real coal mine for simulation testing. After the test, the program is optimized until the crawler vehicle successfully avoids danger. The purpose is to allow the crawler vehicle to adapt to the complex real environment of the coal mine and facilitate the detection of the crawler vehicle driving data under real conditions.

[0041] The specific operation steps of establishing the tracked vehicle obstacle avoidance database in step S1 are as follows:

[0042] S101, collecting a database of tracked vehicle obstacle avoidance, and collecting all tracked vehicle category information and all tracked vehicle information under various categories through a tracked vehicle data center, so as to facilitate rapid search for information categories;

[0043] S102, respectively establish all obstacle avoidance information databases, count the obstacle avoidance information belonging to each category into the corresponding tracked vehicle classification sub-data layer, and form a tracked vehicle information total database to facilitate the collection of obstacle avoidance information;

[0044] S103, recording the first letters of the obstacle avoidance sub-data layer, extracting the first letter of each first letter in the sub-database, and arranging them in alphabetical order, grouping the first letters of each obstacle avoidance database under the same first letter, forming an alphabetical index of the obstacle avoidance database information, which is used to modify and optimize the information in the database, and finally forming a standard obstacle avoidance database, which is convenient for later data modification and optimization.

[0045] The specific operation steps for establishing the basic obstacle avoidance sub-database in step S1 are as follows:

[0046] S111, extracting all obstacle avoidance information from each tracked vehicle database, and recording the first letters of various tracked vehicle obstacle avoidance data, to facilitate data recording and summarization;

[0047] S112, arranging the first letters of the first words of various obstacle avoidance information in alphabetical order, summarizing the first letters of the information in the tracked vehicle data, and creating an information folder under the corresponding obstacle avoidance information to form a basic obstacle avoidance sub-database;

[0048] S113, and using the first English letter of the first word of the information in each obstacle avoidance database to sort in order, forming an index in each obstacle avoidance sub-database to facilitate the subsequent modification of information.

[0049] Wherein, in step S1, the tracked vehicle obstacle avoidance database is interactively connected with each obstacle avoidance basic sub-database via a data hierarchy.

[0050] The specific operation steps of the sensor system identifying the environmental information acquired in real time in step S2 are as follows:

[0051] S21. The sensing system includes an ultrasonic sensor, an infrared sensor, an extended encoder, a visual sensor, a program modifier and a collision sensor, wherein the collision sensor and the extended encoder are both priority sensors, used to handle obstacle avoidance situations; the infrared sensor, the ultrasonic sensor and the visual sensor are all low-priority sensors, which are obstacle avoidance means for detecting distance, path planning and category for the current road section and performing corresponding processing, so as to realize face-to-face and real-time obstacle avoidance operations for the current obstacles; and the program modifier is a necessary means for the tracked vehicle to be unable to avoid obstacles after detection, and the program that cannot avoid obstacles is modified and optimized to achieve a 100% success rate of obstacle avoidance, so as to facilitate comparison and feedback of driving data of the current road section;

[0052] S22. Transmitting data to the control center according to the communication circuit, and then the control center performs hierarchical analysis on the obstacle avoidance data, so as to obtain environmental information in a standard format, so that the control center can use it as a standard for subsequent execution.

[0053] The specific operation steps of the possibility rating and prediction in step S3 are as follows:

[0054] S31. Under the vehicle-mounted system perception system, the matched scene model is compared with the information in the standardized obstacle avoidance database one by one, so as to rate and predict the obstacles encountered by the tracked vehicle during the existing driving process. If the obstacle is huge, multiple routes need to be planned and the travel speed is slow, the rating is S. If there are many obstacles, the routes need to cross and multi-angle adjustment and multi-sensing are required in the middle, the rating is A. If the road is flat during driving and no other emergency measures are required to cross the obstacles, the rating is B, which is convenient for later judgment of the processing hierarchy relationship;

[0055] S32. At the same time, the sensing system inside the crawler vehicle will pre-plan and reorganize the path, making the best choice for the subsequent multiple obstacle sections.

[0056] The specific operation steps of performing the simulation test on the crawler vehicle in step S4 are as follows:

[0057] Combined with the existing obstacle avoidance data information, the tracked vehicle will be driven in a specific environment, and sensors and gravity sensors will be inserted into the obstacle simulators. The number and intensity of the tracked vehicle's collisions with obstacles during driving will be recorded, and the real-time data will be uploaded to the obstacle avoidance big data background. Computer algorithms will be used to compare the multiple collision data before and after, and the tracked vehicle program will be optimized and modified to facilitate the comparison of multiple collision data before and after, and re-optimize the program settings.

[0058] The specific operation steps of performing a real simulation test on the crawler vehicle in step S5 are as follows:

[0059] Based on the number of times the tracked vehicle avoids obstacles and the intensity of impact in the simulated environment, the tracked vehicle is placed in a real coal mine background for simulation testing. The sensing system will upload the real data to the cloud, and the sensing system will plan the travel path and apply it to this driving journey until the tracked vehicle completely avoids obstacles, so that the tracked vehicle can plan a reasonable driving path.

[0060] Specifically: In practical applications, first, a tracked vehicle obstacle avoidance database is established to store the tracked vehicle obstacle avoidance program, and a basic sub-database for obstacle avoidance of the vehicle is established based on data information of other tracked vehicle databases to store obstacle avoidance programs of different tracked vehicles in other environments, and the first letters of the obstacle avoidance sub-data layer are recorded, and the first English letters of each first letter in the sub-database are extracted and arranged in alphabetical order, and the first letters of the English letters in each obstacle avoidance database are attributed to the same first letter to form an alphabetical index of the obstacle avoidance database information, which is used to modify and optimize the information in the database, and finally a standard obstacle avoidance database is formed, and the first English letters of the first letters of the information in each obstacle avoidance database are used to sort them in order, and an index is formed in each obstacle avoidance sub-database;

[0061] When the tracked vehicle is driving in a simulated environment, the on-board system perception system will compare the matched scene model with the information in the standardized obstacle avoidance database one by one, so as to rate and predict the obstacles encountered by the tracked vehicle during the current driving process. If the obstacle is huge, multiple routes need to be planned and the travel speed is slow, the rating is S. If there are many obstacles, the routes need to cross and multi-angle adjustment and multi-sensing are required in the middle, the rating is A. If the road is flat during driving and no other emergency measures are required to climb over the obstacles, the rating is B.

[0062] At the same time, the induction system inside the crawler will pre-plan and reorganize the path, making the best choice for the subsequent multiple obstacle sections;

[0063] Combined with the existing obstacle avoidance data information, the tracked vehicle will be driven in a specific environment, and sensors and gravity sensors will be inserted into the obstacle simulators. The number and strength of the tracked vehicle's collisions with obstacles during driving will be recorded, and the real-time data will be uploaded to the obstacle avoidance big data background in real time. Computer algorithms will be used to compare the multiple collision data before and after, and the tracked vehicle's program will be optimized and modified again. Finally, the test will be carried out again in a real coal mine background, and the crawler vehicle's driving data will be finally obtained. The operator will conduct data analysis and optimization until the tracked vehicle avoids the obstacle completely.

[0064] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for learning obstacle avoidance for a crawler vehicle in a coal mine, characterized in that: The following steps are involved: S1. Establish a tracked vehicle obstacle avoidance database for storing the tracked vehicle obstacle avoidance program, and establish a basic sub-database for the tracked vehicle obstacle avoidance based on the data information of other tracked vehicle databases for storing the obstacle avoidance programs of different tracked vehicles in other environments; S2, the crawler vehicle sensor system identifies the real-time acquired environmental information and transmits it to the control center through the communication circuit. The control center analyzes the sensor data and outputs the environmental information in a standard format; S3, then detect the moving obstacles in the environment during driving, which is completed by the vehicle-mounted environmental perception system, and then rate and predict the possibility of obstacles that may be encountered during driving, judge the collision relationship with obstacles, and make the tracked vehicle avoid obstacles safely through intelligent decision-making and path planning, which is determined by the tracked vehicle path decision system; S4. Simultaneously, a simulation test is performed on the crawler vehicle to simulate the obstacle avoidance path and performance of the crawler vehicle in the coal mine. If the obstacle avoidance effect is not good, the database of the crawler vehicle will be program-changed or optimized, and the crawler vehicle will be simulated again until the obstacle avoidance rate of the crawler vehicle is 100%; The specific operation steps of the simulation test on the crawler vehicle in S4 are: Combined with the existing obstacle avoidance data information, the crawler vehicle will be driven in a simulated real coal mine environment, and sensors and gravity sensors will be inserted into the obstacle simulators. The number and strength of the collisions between the crawler vehicle and the obstacles during driving will be recorded, and the real-time data will be uploaded to the obstacle avoidance big data background in real time. Computer algorithms will be used to compare the multiple collision data before and after, and the crawler vehicle program will be optimized and modified again. S5. Place the crawler vehicle in a real coal mine for simulation testing. After the test, optimize the program until the crawler vehicle successfully avoids danger, so that the crawler vehicle can adapt to the complex real environment of the coal mine.

2. The method for learning obstacle avoidance for a crawler vehicle in an underground coal mine according to claim 1, characterized in that: The specific operation steps of establishing the database for tracked vehicle obstacle avoidance in step S1 are: S101, collecting a database of tracked vehicle obstacle avoidance, and collecting all tracked vehicle category information and all tracked vehicle information under various categories through a tracked vehicle data center; S102, respectively establish all obstacle avoidance information databases, and count the obstacle avoidance information belonging to each category into the corresponding tracked vehicle classification sub-data layer to form a tracked vehicle information total database; S103, recording the first letters of the obstacle avoidance sub-data layer, extracting the first letter of each first letter in the sub-database, and arranging them in alphabetical order, grouping the first letters of each obstacle avoidance database under the same first letter, forming an alphabetical index of the obstacle avoidance database information, which is used to modify and optimize the information in the database, and finally forming a standard obstacle avoidance database.

3. The method for learning obstacle avoidance for a crawler vehicle in an underground coal mine according to claim 2, characterized in that: The specific operation steps of establishing the obstacle avoidance basic sub-database in step S1 are: S111, extracting all obstacle avoidance information from each tracked vehicle database, and recording the first letters of the obstacle avoidance data of each tracked vehicle; S112, arranging the first letters of the first words of various obstacle avoidance information in alphabetical order, summarizing the first letters of the information in the tracked vehicle data, and creating an information folder under the corresponding obstacle avoidance information to form a basic obstacle avoidance sub-database; S113, and use the first English letter of the first word of the information in each obstacle avoidance database to sort them in order, and form an index in each obstacle avoidance sub-database.

4. The method for learning obstacle avoidance for a crawler vehicle in an underground coal mine according to claim 1, characterized in that: In step S1, the tracked vehicle obstacle avoidance database is interactively connected with each obstacle avoidance basic sub-database through a data hierarchy.

5. The method for learning obstacle avoidance for a crawler vehicle in an underground coal mine according to claim 1, characterized in that: The specific operation steps of the sensor system identifying the environmental information acquired in real time in step S2 are: S21. The sensing system includes an ultrasonic sensor, an infrared sensor, an extended encoder, a visual sensor, a program modifier and a collision sensor, wherein the collision sensor and the extended encoder are both priority sensors used to handle obstacle avoidance situations; the infrared sensor, the ultrasonic sensor and the visual sensor are all low-priority sensors, which are obstacle avoidance means for detecting the distance and path planning and category of the current road section and performing corresponding processing, so as to realize the step of face-to-face and real-time obstacle avoidance operation for the current obstacle; and the program modifier is a necessary means for the tracked vehicle to be unable to avoid obstacles after detection, and the program that cannot avoid obstacles is modified and optimized to achieve a 100% success rate of obstacle avoidance; S22. Transmitting data to the control center according to the communication circuit, and then the control center performs hierarchical analysis on the obstacle avoidance data, so as to obtain environmental information in a standard format, so that the control center can use it as a standard for subsequent execution.

6. The method for learning obstacle avoidance for a crawler vehicle in an underground coal mine according to claim 5, characterized in that: The specific operation steps of the possibility rating and prediction in step S3 are: S31. Under the vehicle-mounted system perception system, the matched scene model is compared with the information in the standardized obstacle avoidance database one by one, so as to rate and predict the obstacles encountered by the tracked vehicle during the current driving process. If the obstacle is huge, multiple routes need to be planned and the travel speed is slow, the rating is S. If there are many obstacles, the routes need to cross and multi-angle adjustment and multi-sensing are required in the middle, the rating is A. If the road is flat during driving and no other emergency measures are required to climb over the obstacles, the rating is B. S32. At the same time, the sensing system inside the tracked vehicle will pre-plan and reorganize the path, making the best choice for the subsequent multiple obstacle sections.

7. The method for learning obstacle avoidance for a crawler vehicle in an underground coal mine according to claim 1, characterized in that: The specific operation steps of performing a real simulation test on the crawler vehicle in step S5 are: Based on the number of times the tracked vehicle avoids obstacles and the intensity of impact in the simulated environment, the tracked vehicle is placed in a real coal mine background for simulation testing. The sensing system will upload the real data to the cloud, and the sensing system will plan the travel path and apply it to this driving journey until the tracked vehicle avoids the obstacle and passes it completely.

Images