阅读说明：本技术 一种移动机器人在核事故复杂环境中的自主导航方法 (Autonomous navigation method of mobile robot in nuclear accident complex environment ) 是由 熊振华 黄颖 周研 吴建华 于 2021-07-15 设计创作，主要内容包括：本发明公开了一种移动机器人在核事故复杂环境中的自主导航方法,涉及核工程与核技术领域,该方法包括如下步骤：对辐射场分布地图进行处理,进行栅格化,得到辐射代价地图；对环境点云地图进行处理,投影成灰度图像,处理获得图像梯度,得到坡度代价地图；将辐射代价地图与坡度代价地图叠加,得到多层代价地图；根据环境点云地图以及当前雷达数据,利用Scan Context算法进行定位,获得机器人在环境中的位置；根据多层代价地图和定位信息,使用改进A*算法规划合理路径,使机器人到达目标位置。本发明可以在核事故复杂环境下同时考虑多种因素,为机器人规划安全高效的路径,实现安全可靠的自主导航。(The invention discloses an autonomous navigation method of a mobile robot in a complex nuclear accident environment, which relates to the technical field of nuclear engineering and nuclear, and comprises the following steps: processing the radiation field distribution map, and rasterizing to obtain a radiation cost map; processing the environmental point cloud map, projecting the environmental point cloud map into a gray image, and processing the gray image to obtain an image gradient and obtain a gradient cost map; overlapping the radiation cost map and the gradient cost map to obtain a multilayer cost map; positioning by utilizing a Scan Context algorithm according to the environment point cloud map and the current radar data to obtain the position of the robot in the environment; and planning a reasonable path by using an improved A-algorithm according to the multilayer cost map and the positioning information, so that the robot reaches a target position. The invention can simultaneously consider a plurality of factors under the complex environment of the nuclear accident, plan a safe and efficient path for the robot and realize safe and reliable autonomous navigation.)

1. An autonomous navigation method of a mobile robot in a complex environment of nuclear accidents, characterized in that the method comprises the following steps:

step 1, processing a radiation field distribution map, and rasterizing to obtain a radiation cost map;

step 2, processing the environmental point cloud map, projecting the environmental point cloud map into a gray image, and processing the gray image to obtain an image gradient and obtain a gradient cost map;

step 3, overlapping the radiation cost map and the gradient cost map to obtain a multilayer cost map;

step 4, positioning by utilizing a Scan Context algorithm according to the environment point cloud map and the current radar data to obtain positioning information of the robot in the environment;

and 5, planning a reasonable path by using an improved A-star algorithm according to the multilayer cost map and the positioning information, so that the robot reaches a target position.

2. The method for autonomous navigation of a mobile robot in a complex environment of nuclear accidents according to claim 1, wherein said radiation cost map comprises a radiation alert cost map and a radiation dose cost map, and said grade cost map comprises a grade alert cost map and a grade mobile cost map.

3. The method for autonomous navigation of a mobile robot in a complex environment of nuclear accidents according to claim 2, wherein in step 1, the radiation field distribution mapping process specifically comprises the following steps:

step 1.1, selecting a proper grid size;

step 1.2, calculating the radiation dose rate of the corresponding position of each grid to obtain a radiation dose cost map;

and 1.3, selecting a radiation dose threshold value, and marking the grid with the radiation dose exceeding the radiation dose threshold value as an obstacle to obtain the radiation warning cost map.

4. The autonomous navigation method of the mobile robot in the complex environment of nuclear accident according to claim 2, wherein in the step 2, the environmental point cloud map processing process specifically includes the following steps:

step 2.1, cutting the point cloud, and removing invalid information of the high ceiling in the path planning;

2.2, identifying the ground by using a RANSAC algorithm;

step 2.3, correcting the point cloud, and rotating the point cloud to enable the ground to be coincident with a horizontal plane;

2.4, selecting a proper grid size;

step 2.5, calculating the average value of the height of the point cloud in the corresponding area of the grid; linearly mapping the point cloud to a gray value of a gray image according to the maximum value and the minimum value of the height of the point cloud, and projecting the point cloud into the gray image;

step 2.6, performing closed operation on the gray level image to fill in holes;

step 2.7, calculating gradients in the directions of 0 degree, 45 degrees, 90 degrees and 135 degrees by using a sobel operator to obtain 4 gradient moving cost maps;

and 2.8, selecting a gradient threshold value, and marking the grid with the gradient exceeding the gradient threshold value as an obstacle to obtain the gradient warning cost map.

5. The autonomous navigation method of a mobile robot in a complex nuclear accident environment according to claim 2, wherein in the step 3, the multi-layer cost map processing procedure specifically includes the following steps:

step 3.1, superposing the radiation warning cost map and the gradient warning cost map, and marking the grids with the radiation dose exceeding the radiation dose threshold or the gradient exceeding the gradient threshold as obstacles;

and 3.2, storing the radiation dose cost map and the gradient moving cost map separately, and storing the corresponding cost moved to the grid in the grid.

6. The method for autonomous navigation of a mobile robot in a complex environment of nuclear accidents according to claim 1, wherein in the step 4, the Scan Context algorithm specifically comprises the following steps:

step 4.1, dividing the point cloud of the current radar data into 20 rings, wherein each ring is divided into 60 equal parts, and 1200 grids are obtained;

4.2, calculating the maximum height value of the point cloud in each grid, and representing the point cloud by using a two-dimensional image scan context;

4.3, further calculating the mean value of the columns to obtain a 1 multiplied by 60 column vector ring key; calculating the mean value of the lines to obtain a 20 multiplied by 1 line vector key;

step 4.4, constructing a kd-tree by using the column vector ring key, and executing KNN search to obtain a candidate matching scan context;

step 4.5, aligning the candidate matching scan context by using the row vector sector key to obtain an offset;

and 4.6, applying an offset to the candidate matching scan context, and then comparing to obtain the position of the robot.

7. The autonomous navigation method of a mobile robot in a complex environment of nuclear accident according to claim 2, characterized in that in said step 5, said improvement a algorithm constructs a path evaluation function f (n) ═ g (n) + h (n), g (n) is the actual cost spent from the starting node to the current node n, h (n) is the estimated cost from the current node n to the target node; g (n) and H (n) are related to the length of the path, the radiation dose and the road surface gradient.

8. The method for autonomous navigation of a mobile robot in a complex environment of nuclear accidents according to claim 2, characterized in that in said step 5, said improvement a algorithm comprises in particular the following steps:

step 5.1, setting a target node, wherein an initial node is the current position of the robot;

step 5.2, two lists are created: an open list and a close list;

step 5.3, adding the starting node into the open list;

step 5.4, traversing the nodes in the open list, when the open list is empty, indicating that no path is found, ending the algorithm, otherwise, turning to the next step;

step 5.5, when the open list is not empty, selecting a node with the minimum path cost function value from the open list, and adding the node into the closed list;

step 5.6, judging whether the current node is the target node; if the current node is the target node, searching is finished, and the optimal path is found, otherwise, the step 5.7 is carried out;

step 5.7, expanding the current node, and calculating the path cost function values of all the child nodes, wherein the step specifically comprises the following steps:

step 5.7.1, if the child node is not in the open list and the closed list, adding the child node to the open list, and adding a pointer to the current node to the child node; after the path search is finished, connecting the paths according to the pointing direction of the pointer;

step 5.7.2, if the child node is already in the open list, comparing the new function value obtained by calculation with the old value in the open list, saving the node with smaller function value, and modifying the pointer to point to the current child node;

step 5.7.3, if the child node is in the closed list, ignoring the node, and returning to the step 5.7 to continue expanding other nodes;

and 5.8, finishing the algorithm when the optimal solution or no solution is found, and returning to the step 5.4 to continue circulation if the optimal solution or no solution is found.

9. The method for autonomous navigation of a mobile robot in a complex environment of nuclear accidents according to claim 3, characterized in that in step 1, the cumulative radiation dose of the robot along the path from the starting point to the end point is related to the dose rate of each grid, the motion speed of the robot and the grid size.

10. The autonomous navigation method of a mobile robot in a nuclear accident complex environment according to claim 4, wherein in step 2, the sobel operator is a gradient operator matrix of 3 x 3, and the sobel operator is convolved with an image to obtain 4 gradient mobile cost maps.

Technical Field

The invention relates to the technical field of nuclear engineering and nuclear, in particular to an autonomous navigation method of a mobile robot in a nuclear accident complex environment.

Background

With the rapid development of the nuclear industry and the wide application of nuclear power technology, the need of people for the safety guarantee of nuclear radiation is increasingly urgent. In the daily operation of the nuclear power station, workers are inevitably affected by radiation, and personal safety is endangered. In recent decades, the increasing development of robotics has made it possible for robots to perform work in a radiation environment. Most of nuclear environment robots developed at the present stage are mobile remote-controlled robots, but remote control operation is difficult under many emergency situations. This puts new demands on the robot for safe and autonomous navigation in a radiation environment.

First, a strong radiation environment can cause the electronics in the sensor to malfunction, causing the robot to be lost in the radiation environment. However, due to technical and cost limitations, the radiation-resistant design of the robot still cannot be made to be capable of disregarding the radiation dose. This requires that the robot avoid radiation as much as possible while navigating autonomously. Secondly, in an accident environment, obstacles like collapsed walls may appear, constituting a slope on the ground. At present, the mobile robot navigation generally uses a two-dimensional map, and the obstacles can be classified into impassable areas. But in reality, the emergency robot has certain climbing capacity and can pass through obstacles with small gradient. If the grade effect is not considered, path planning may fail.

The patent application No. 202010416736.2 discloses a rescue robot for nuclear radiation environment, which includes a path planning system, and can plan the traveling route of the robot body by analyzing the radiation signal and the environment signal. A radiation signal threshold value, a target distance threshold value and a distance threshold value between the target and the obstacle are preset, and the continuation or the replanning is selected by comparing with the preset threshold value.

Although the technical scheme disclosed in the patent proposes a self-operation control method of the robot, the method is not reliable enough and is not suitable for complex nuclear accident environment. The feasibility of the environment cannot be accurately identified, and the overall cost of the path cannot be estimated.

Therefore, those skilled in the art are dedicated to develop an autonomous navigation method of a mobile robot in a complex environment of a nuclear accident, and plan a safe and efficient path for the robot by considering various factors in the complex environment.

Disclosure of Invention

In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is how to accurately describe the radiation characteristics and physical trafficability characteristics of the environment, and to plan the most suitable path for the robot in a complex environment by comprehensively considering the path length, the radiation dose and the road surface gradient.

In order to achieve the purpose, the invention provides an autonomous navigation method of a mobile robot in a complex environment of a nuclear accident, which comprises the following steps:

step 1, processing a radiation field distribution map, and rasterizing to obtain a radiation cost map;

step 2, processing the environmental point cloud map, projecting the environmental point cloud map into a gray image, and processing the gray image to obtain an image gradient and obtain a gradient cost map;

step 3, overlapping the radiation cost map and the gradient cost map to obtain a multilayer cost map;

step 4, positioning by utilizing a Scan Context algorithm according to the environment point cloud map and the current radar data to obtain positioning information of the robot in the environment;

and 5, planning a reasonable path by using an improved A-star algorithm according to the multilayer cost map and the positioning information, so that the robot reaches a target position.

Further, the radiation cost map comprises a radiation warning cost map and a radiation dose cost map, and the gradient cost map comprises a gradient warning cost map and a gradient mobile cost map.

Further, in step 1, the radiation field distribution map processing process specifically includes the following steps:

step 1.1, selecting a proper grid size;

step 1.2, calculating the radiation dose rate of the corresponding position of each grid to obtain a radiation dose cost map;

and step 1.3, selecting a radiation dose threshold value, and marking the grid with the radiation dose exceeding the radiation dose threshold value as an obstacle to obtain a radiation warning cost map.

Further, in step 2, the processing process of the environment point cloud map specifically includes the following steps:

step 2.1, cutting the point cloud, and removing invalid information of the high ceiling in the path planning;

2.2, identifying the ground by using a RANSAC algorithm;

step 2.3, correcting the point cloud, and rotating the point cloud to enable the ground to coincide with the horizontal plane;

2.4, selecting a proper grid size;

step 2.5, calculating the average value of the height of the point cloud in the corresponding area of the grid; linearly mapping the point cloud to the gray value of the gray image according to the maximum value and the minimum value of the height of the point cloud, and projecting the point cloud into the gray image;

step 2.6, performing closed operation on the gray level image to fill the hole;

step 2.7, calculating gradients in the directions of 0 degree, 45 degrees, 90 degrees and 135 degrees by using a sobel operator to obtain 4 gradient moving cost maps;

and 2.8, selecting a gradient threshold value, and marking the grid with the gradient exceeding the gradient threshold value as an obstacle to obtain a gradient warning cost map.

Further, in step 3, the multi-layer cost map processing process specifically includes the following steps:

step 3.1, superposing the radiation warning cost map and the gradient warning cost map, and marking the grids with the radiation dose exceeding a radiation dose threshold or the gradient exceeding a gradient threshold as obstacles;

and 3.2, storing the radiation dose cost map and the gradient moving cost map separately, and storing the corresponding cost moved to the grid in the grid.

Further, in step 4, the Scan Context algorithm specifically includes the following steps:

step 4.1, dividing the point cloud of the current radar data into 20 rings, wherein each ring is divided into 60 equal parts, and 1200 grids are obtained;

4.2, calculating the maximum height value of the point cloud in each grid, and representing the point cloud by using a two-dimensional image scan context;

4.3, further calculating the mean value of the columns to obtain a 1 multiplied by 60 column vector ring key; calculating the mean value of the lines to obtain a 20 multiplied by 1 line vector key;

step 4.4, constructing a kd-tree by using the column vector ring key, and executing KNN search to obtain a candidate matching scan context;

step 4.5, aligning the candidate matching scan context by using a row vector sector key to obtain an offset;

and 4.6, applying an offset to the candidate matching scan context, and then comparing to obtain the position of the robot.

Further, in step 5, the improved a algorithm constructs a path evaluation function f (n) ═ g (n) + h (n), g (n) is the actual cost spent from the starting node to the current node n, and h (n) is the estimated cost from the current node n to the target node; g (n) and H (n) are related to the length of the path, the radiation dose and the road surface gradient.

Further, in step 5, the improved a-algorithm specifically includes the following steps:

step 5.1, setting a target node, wherein an initial node is the current position of the robot;

step 5.2, two lists are created: an open list and a close list;

step 5.3, adding the starting node into the open list;

step 5.4, traversing nodes in the open list, when the open list is empty, indicating that no path is found, ending the algorithm, otherwise, turning to the next step;

step 5.5, when the open list is not empty, selecting a node with the minimum path cost function value from the open list, and adding the node into the closed list;

step 5.6, judging whether the current node is a target node; if the current node is the target node, the searching is finished, the optimal path is found, otherwise, the step 5.7 is carried out;

step 5.7, expanding the current node, and calculating the path cost function values of all the sub-nodes, wherein the step specifically comprises the following steps;

step 5.7.1, if the child node is not in the open list and the closed list, adding the child node into the open list, and adding a pointer pointing to the current node to the child node; after the path search is finished, connecting the paths according to the pointing direction of the pointer;

step 5.7.2, if the child node is already in the open list, comparing the new function value obtained by calculation with the old value in the open list, saving the node with smaller function value, and modifying the pointer to point to the current child node;

step 5.7.3, if the child node is in the closed list, ignoring the node, and returning to the step 5.7 to continue expanding other nodes;

and 5.8, finishing the algorithm when the optimal solution or no solution is found, and returning to the step 5.4 to continue circulation if the optimal solution or no solution is found.

Further, in step 1, the cumulative radiation dose of the robot along the path from the start point to the end point is related to the dose rate of each grid, the motion speed of the robot and the grid size.

Further, in step 2, the sobel operator is a gradient operator matrix of 3 × 3, and the sobel operator is convolved with the image to obtain 4 gradient moving cost maps.

Compared with the prior art, the invention at least has the following beneficial technical effects:

according to the invention, the multi-layer cost map is used, the environmental information can be better represented, the trafficability and traffic cost of the whole environment can be rapidly obtained by processing the radiation field distribution map and the environmental point cloud map, and an accurate and reliable obstacle map and cost map are provided for the following improved A-ray algorithm path planning; when the environment point cloud map is processed, the point cloud is projected into a gray image, and an image processing method is used for replacing the point cloud processing method; noise can be removed through image operation, grids without information are filled, and gradient information corresponding to the ground gradient can be rapidly calculated; an improved A-star algorithm is used, the response to the environment is rapid, and the heuristic information is used for searching the optimal path; the heuristic function comprehensively considers information such as path length, radiation dose, road surface gradient and the like, and can complete path planning under different task requirements by modifying the influence coefficient and searching the optimal path under specific conditions.

The conception, the specific structure and the technical effects of the present invention will be further described with reference to the accompanying drawings to fully understand the objects, the features and the effects of the present invention.

Drawings

FIG. 1 is a flow chart of an autonomous navigation method of a mobile robot in a complex environment of a nuclear accident according to a preferred embodiment of the present invention;

FIG. 2 is a schematic diagram of the multi-level cost map structure organization of a preferred embodiment of the present invention;

fig. 3 is a schematic diagram of an improved a algorithm flow according to a preferred embodiment of the present invention.

Detailed Description

The technical contents of the preferred embodiments of the present invention will be more clearly and easily understood by referring to the drawings attached to the specification. The present invention may be embodied in many different forms of embodiments and the scope of the invention is not limited to the embodiments set forth herein.

In the drawings, structurally identical elements are represented by like reference numerals, and structurally or functionally similar elements are represented by like reference numerals throughout the several views. The size and thickness of each component shown in the drawings are arbitrarily illustrated, and the present invention is not limited to the size and thickness of each component. The thickness of the components may be exaggerated where appropriate in the figures to improve clarity.

As shown in fig. 1, the method for autonomous navigation of a mobile robot in a complex environment of a nuclear accident according to a preferred embodiment of the present invention includes the following steps:

step 1, processing a radiation field distribution map, and rasterizing to obtain a radiation cost map;

step 2, processing the environmental point cloud map, projecting the environmental point cloud map into a gray image, and processing the gray image to obtain an image gradient and obtain a gradient cost map;

step 3, overlapping the radiation cost map and the gradient cost map to obtain a multilayer cost map;

step 4, positioning by utilizing a Scan Context algorithm according to the environment point cloud map and the current radar data to obtain the position of the robot in the environment;

and 5, planning a reasonable path by using an improved A-star algorithm according to the multilayer cost map and the positioning information, so that the robot reaches a target position.

As shown in fig. 2, which is a multi-layer cost map structure according to a preferred embodiment of the present invention, the radiation cost map includes a radiation warning cost map and a radiation dose cost map, and the gradient cost map includes a gradient warning cost map and a gradient moving cost map.

Specifically, the specific steps of processing and obtaining the radiation cost map include:

step 1.1, selecting a proper grid size, rasterizing a radiation field distribution map, and storing the average radiation dose rate of a corresponding position in each grid;

step 1.2, selecting the maximum instantaneous dose rate which can be borne by the robot as a threshold, and when the radiation dose rate in the environment is greater than the threshold, considering that a fatal obstacle exists at the position, thereby obtaining a radiation alert cost map. Stored value c in radiation warning cost map_1[index]The following were used:

step 1.3, the cumulative radiation dose of the robot along the path P from the start point S to the end point G is related to the dose rate per grid R (i, j), the motion speed v of the robot and the grid size Δ S. The total radiation E received by the robot is calculated as follows:

to simplify the calculation of the path evaluation function, the value c is stored in a radiation dose cost map_2[index]The following were used:

specifically, the concrete steps of processing and obtaining the gradient cost map include:

step 2.1, cutting the point cloud, and removing invalid information of high ceilings and the like in path planning by setting the maximum value of a Z axis of the point cloud;

2.2, identifying the ground by using a RANSAC algorithm to obtain plane parameters;

step 2.3, correcting the point cloud to enable the ground normal vector to coincide with the Z axis, and calculating a rotation matrix according to the ground normal vector and the Z axis by a Rodrigue rotation formula to enable the ground to coincide with the horizontal plane;

2.4, selecting a proper grid size;

step 2.5, calculating the average value of the height of the point cloud in the corresponding area of the grid, linearly mapping the point cloud to the gray value of the gray image according to the maximum value and the minimum value of the height of the point cloud, wherein the range is 0-255, and projecting the point cloud into the gray image;

step 2.6, performing closed operation on the gray level image to fill the hole;

and 2.7, calculating gradients in the directions of 0 degrees, 45 degrees, 90 degrees and 135 degrees by using a sobel operator. The sobel operator is a gradient operator matrix of 3 multiplied by 3, and the sobel operator and the image are convoluted to obtain 4 gradient moving cost maps. The gradient values in the directions of 0 degree, 45 degrees, 90 degrees and 135 degrees are respectivelyStored value c in radiation dose cost map_4[index]The following were used:

and 2.8, selecting the maximum slope which can be crossed by the robot as a threshold, and marking the grid with the slope exceeding the threshold as an obstacle to obtain a slope warning cost map. Stored value c in slope warning cost map_3[index]The following were used:

specifically, the multilayer cost map processing process specifically includes the steps of:

and overlapping the radiation cost map and the gradient cost map to obtain a multilayer cost map. And marking the grids with the radiation dose exceeding the threshold or the gradient exceeding the threshold as the obstacles to obtain an obstacle map. The radiation dose cost map and the gradient moving cost map are stored separately, and the radiation dose cost map and the gradient moving cost map are convenient to use when a path evaluation function is calculated by a follow-up improved A-star algorithm.

Specifically, according to an environment point cloud map and current radar data, a Scan Context algorithm is used for positioning, and the method specifically comprises the following steps:

and 4.1, dividing the point cloud of the current radar data into 20 rings, wherein each ring is divided into 60 equal parts, and the total number of the rings is 1200 grids.

And 4.2, calculating the maximum height value of the point cloud in each grid, representing the point cloud by using a two-dimensional image, and marking as scan context.

4.3, further calculating the mean value of the rows to obtain a vector of 1 multiplied by 60, and recording the vector as a ring key; the mean of the rows is calculated to obtain a 20 × 1 vector, which is denoted as sector key.

And 4.4, constructing a kd-tree by using the ring key, and executing knn search to obtain a candidate matching scan context.

And 4.5, aligning the candidate matching scan context by using sector key to obtain the offset.

And 4.6, applying an offset to the candidate matching scan context, and then comparing to obtain the position of the robot.

Specifically, a reasonable path is planned by using an improved A-star algorithm according to the multilayer cost map and the positioning information, so that the robot reaches a target position. The path evaluation function f (n) ═ g (n) + h (n), g (n) represents the actual cost spent from the start node to the current node n, and h (n) is the estimated cost from the current node n to the target node.

As shown in fig. 3, the specific steps of the improved a-algorithm are as follows:

step 5.1, setting a target node, wherein an initial node is the current position of the robot;

step 5.2, two lists are created: an open list and a close list;

step 5.3, adding the starting node into the open list;

step 5.4, traversing nodes in the open list, when the open list is empty, indicating that no path is found, ending the algorithm, otherwise, turning to the next step;

step 5.5, when the open list is not empty, selecting a node with the minimum path cost function value from the open list, and adding the node into the closed list;

step 5.6, judging whether the current node is a target node; if the current node is the target node, the searching is finished, the optimal path is found, otherwise, the step 5.7 is carried out;

step 5.7, expanding the current node and calculating the path cost function values of all the sub-nodes;

step 5.7.1, if the child node is not in the open list and the closed list, adding the child node into the open list, and adding a pointer pointing to the current node to the child node; after the path search is finished, connecting the paths according to the pointing direction of the pointer;

step 5.7.2, if the child node is already in the open list, comparing the new function value obtained by calculation with the old value in the open list, saving the node with smaller function value, and modifying the pointer to point to the current child node;

step 5.7.3, if the child node is in the closed list, ignoring the node, and returning to the step 5.7 to continue expanding other nodes;

and 5.8, finishing the algorithm when the optimal solution or no solution is found, and returning to the step 5.4 to continue circulation if the optimal solution or no solution is found.

Specifically, in step 5.7, the step of calculating the path evaluation function specifically includes:

1. calculation of the actual cost g (n).

Path factor x_kThe definition is as follows:

the calculation formula of the path length L is as follows:

the calculation formula of the total irradiation quantity of the robot is as follows:

the total gradient cost calculation formula of the robot movement is as follows:

introducing a cost influence coefficient epsilon₁,ε₂And ε₃The impact of path, radiation and grade costs on the total cost is represented separately. Normalizing each part of cost, wherein the calculation formula of the actual cost G (n) is as follows:

2. calculation of the estimated cost H (n).

Let dx and dy denote the horizontal and vertical distances between the current node N and the target node G. The shortest distance to the target node is calculated as follows:

the minimum number of moves to the target node is calculated as follows:

N_min＝max(dx,dy)

the calculation formula of the average radiation dose of the rectangular area with the former node N and the target node G as vertexes is as follows:

the calculation formula of the average gradient of the rectangular area with the former node N and the target node G as vertexes is as follows:

normalizing each part of the cost, and calculating the estimated cost H (n) according to the following formula:

3. and calculating a path evaluation function. The calculation formula of the path evaluation function f (n) is as follows:

F(n)＝G(n)+H(n)

the foregoing detailed description of the preferred embodiments of the invention has been presented. It should be understood that numerous modifications and variations could be devised by those skilled in the art in light of the present teachings without departing from the inventive concepts. Therefore, the technical solutions available to those skilled in the art through logic analysis, reasoning and limited experiments based on the prior art according to the concept of the present invention should be within the scope of protection defined by the claims.