^{1,2}

^{1,2}

^{1}

^{2}

As the explosion-proof safety level of a coal mine robot has not yet reached the level of intrinsic safety “ia” and it cannot work in a dangerous gas distribution area, therefore, path planning methods for coal mine robot to avoid the dangerous area of gas are necessary. In this paper, to avoid a secondary explosion when the coal mine robot passes through gas hazard zones, a path planning method is proposed with consideration of gas concentration distributions. First, with consideration of gas distribution area and obstacles, MAKLINK method is adopted to describe the working environment network diagram of the coal mine robot. Second, the initial working paths for the coal mine robot are obtained based on Dijkstra algorithm, and then the global optimal working path for the coal mine robot is obtained based on ant colony algorithm. Lastly, experiments are conducted in a roadway after an accident, and results by different path planning methods are compared, which verified the effectiveness of the proposed path planning method.

Environment exploration (such as detection of methane concentration) is often necessary before coal mining, and a coal mine robot is needed. Similarly, rescue must be carried out after a coal mine accident, and the working environment in the roadway is always very complex; therefore, coal mine robot is also a good choice. In the process of environment detection or rescue, a secondary explosion must be avoided.

In China, the coal mine safe electrical apparatus is Class I, which is strict with the use of mining equipment. Therefore, with regard to coal mine robot, the explosion-proof grade needs to reach the level of intrinsic safety “ia.” However, at present, very few apparatuses can reach the level of “ia” [

Path planning, which can find out the optimal working path without obstacles, is very important for coal mine robot. Generally, path planning is a nondeterministic polynomial problem with such constraints from much information of environment and obstacles [

In the paper, to avoid a secondary explosion when the coal mine robot passes through gas hazard zones, path planning method is studied. First, MAKLINK method is adopted to describe the working environment network diagram of the coal mine robot with consideration of gas distribution area and obstacles. Second, with the help of Dijkstra algorithm, the initial working paths for coal mine robot are obtained, and then ant colony algorithm is adopted to optimize the initial working paths, and the global optimal working paths for coal mine robot are obtained. Lastly, experiments are conducted in a roadway after an accident, and results by different path planning methods are compared.

In this paper, with consideration of gas distribution area and obstacles, MAKLINK method is adopted to describe the working environment network graph of coal mine robot. Suppose that the roadway bottom is parallel to

Based on the GIS system before the mine accident and the modeling of roadway environment after the mine accident, the optimal working path for coal mine robot is obtained by Dijkstra-ant colony algorithm. The optimal working path planning method for coal mine robot is shown in Figure

Dijkstra-ant colony algorithm for coal mine robot.

The initial working paths for coal mine robot are obtained based on Dijkstra algorithm. The working environment network graph obtained by MAKLINK method is defined as

The adjacency matrix of the network graph

The set of path nodes of coal mine robot is divided into two subsets

Based on Dijkstra algorithm, the steps to obtain the initial working paths for coal mine robot are as follows:

Put the starting point St into

Select the minimum value

Use point

Repeat step (2) and step (3), until the set

As it is known that a safe working path for coal mine robot can be determined through Dijkstra algorithm, due to the inherent shortcomings of Dijkstra algorithm, it is difficult to guarantee that the obtained path is the optimal one. Therefore, to get the optimal working path for coal mine robot, the initial working paths obtained by Dijkstra algorithm are further optimized by ant colony algorithm.

A working path from

The initial length of the paths and the distances between the path nodes and the nodes of obstacles or gas concentration distributions should be considered simultaneously when ant colony algorithm is used. The objective function of the optimal working path for coal mine robot is defined as follows [

The edges on the initial path in MAKLINK graph are defined as

For a given set of parameters

The initial path

There are

During ant colony algorithm solving process, to avoid falling into a local optimum, a randomly selected parameter

After the selective probability of the

After all the

By using ant colony algorithm, the steps of the working path optimization for the rescue robot can be described as follows [

Get the initial path by Dijkstra algorithm, and initialize parameters of ant colony algorithm.

Calculate the selective probability of each ant, and employ a roulette method to determine which node the ant will move to, and move the ant.

Update the pheromone concentration.

Repeat step (2) and step (3) until all the ants arrive at

Compute the objective functions

If the loop ends, export the optimal path; otherwise, go to step (2).

The proposed path planning method is suitable for the coal mine robot that works in the roadway of longwall mine.

The experiments are carried out in the roadway of Huainan Panyi Mine, whose length is 50 meters and width is 3 meters. In the roadway, gas is distributed regionally and the gas concentration is not over the standard (≤5%). To ensure the safety of the experiments, the area is assumed to be a gas hazard area if the area gas concentration is greater than 2%. Additionally, the length and width of the coal mine robot are 600 mm and 450 mm, respectively. The picture of the coal mine robot is shown in Figure

The coal mine robot.

The data collected by the coal mine robot is processed, and the gas concentration distributions are shown in Figure

Gas concentration distributions.

Regional distributions of gas prone to explosion.

The proposed method is applied to the working path planning for the coal mine robot with only the gas concentration area distributions. The coordinate of the starting point and the target point for the rescue robot is (1.5 m, 0) and (1 m, 50 m), respectively. Dijkstra algorithm and Dijkstra-ant colony optimization algorithm are used to solve the optimal path planning of coal mine robot. The parameters of ant colony algorithm are listed as follows:

Result of the path planning with only consideration of gas concentration distributions.

Result of the path planning

Variation of the total path length

As observed from Figure

The working environment in the roadway is always very complex after a coal mine accident. The proposed method is also applied to the working path planning for the coal mine robot with consideration of the gas concentration distributions and obstacles simultaneously. The final results of the path planning are shown in Figure

Result of the path planning with consideration of gas concentration distributions and obstacles.

Result of the path planning

Variation of the total path length

As observed from Figure

From Figures

To avoid a secondary explosion when the coal mine robot passes through gas hazard zones, a path planning method is proposed with consideration of gas concentration distributions and obstacles simultaneously. First, MAKLINK method is adopted to describe the working environment network diagram of the coal mine robot. Second, the initial working paths for the coal mine robot are obtained based on Dijkstra algorithm, and then the global optimal working path for the coal mine robot is obtained based on ant colony algorithm. To verify the effectiveness of the proposed method, lastly, experiments are conducted in a roadway after an accident, and results by different path planning methods are compared. The experiments are divided into two cases; in the first case, it is assumed that the roadway is a flat roadway, and the robot can gain access through the tunnel except the side wall and the dangerous area of the gas; in the other case, obstacles exist in the roadway, such as stones, mound, and gas danger area.

With only consideration of gas concentration distributions, the total path length by Dijkstra algorithm is 63.39 m, while the one by the proposed method is 60.71 m. With consideration of gas concentration distributions and obstacles, the total path length by Dijkstra algorithm is 63.39 m, while the one by the proposed method is 60.71 m. From the results of the path planning with only consideration of gas concentration distributions and the path planning with consideration of gas concentration distributions and obstacles, it can be seen that the planning paths through the proposed method are shorter, which is of great significance to the battery powered coal mine robot. Therefore, the proposed path planning method possesses better performance. The experimental results verified the effectiveness of the proposed path planning method.

The authors declare that they have no competing interests.

This work is supported by a grant of the National 863 Program of China (no. 2012AA041504).