CN104914219B - The robot gas leakage source localization method long-pending based on matrix semi-tensor and system - Google Patents

Abstract

The robot gas leakage source localization method long-pending based on matrix semi-tensor and system, method comprises the following steps: step S1: determine the robot number participating in gas leakage source location, and the input variable of fuzzy control rule and output variable, input variable is the sensor information that robot detects in real time, and output variable is location behavior; Step S2: set up the input variable fuzzy rule base to output variable; Step S3: utilize matrix semi-tensor long-pending theory that fuzzy control rule is converted into structure matrix; Step S4: choose the behavior of corresponding location according to the sensor information that robot detects in real time, thus completing the work of gas leakage source location. The present invention has merged the long-pending theory of matrix semi-tensor and fuzzy control theory, adopt multiple adjustment modes and control strategy, have accurate positioning, pardon is strong, safety is high, motility is strong and the feature such as practical, meets the gas leakage source location requirement under varying environment.

Description

Robot gas leakage source positioning method and system based on matrix half-tensor product

Technical Field

The invention relates to a gas leakage source positioning method and system, in particular to a robot gas leakage source positioning method and system based on matrix half tensor products, and belongs to the technical field of gas leakage source positioning.

Background

The positioning of the dangerous gas leakage source has very important significance on human safety, such as ocean and river environment monitoring, pollutant source tracking and positioning, toxic gas leakage source in a chemical plant and the like. Therefore, how to quickly and effectively locate the source of a dangerous gas leak is an extremely important issue. However, the problem of locating the source of the hazardous gas leak presents different characteristics in different environments. Generally, in the absence of an air flow, diffusion of the scent molecules is a major force that can drive the scent molecules away from the scent source, where the greatest concentration will occur. Thus, we can use a gradient approach to locate the scent source. However, in the real world, airflow is a major force affecting the diffusion of odors, which forms plume by affecting the movement of odor molecules.

The robot is mainly divided into three stages for positioning the gas leakage source, namely a smoke plume finding stage, a smoke plume tracking stage and an odor source confirming stage. In an air flow environment where the actual wind speed/direction change is relatively large, the gas concentration tends to be meandering or discontinuous due to the influence of turbulence. In addition, due to the limitation of the detection distance and sensitivity of the gas sensor, the robot cannot detect gas concentration information through smell at first in most cases, and smoke plume discovery is mainly carried out in a random search mode, namely the robot adopts the random search mode, the probability of gas leakage source distribution in a search scene is considered to be equal, and then the scene is traversed, so that the search mode is low in efficiency and has certain blindness; in the smoke plume tracking process (tracking the gas plume according to concentration information or/and wind speed and direction information to approach a gas leakage source), a plurality of feasible paths with large differences can be obtained according to smell information, and the robot faces the difficulty of path selection; in the process of confirming the gas leakage source, a concentration method is mainly adopted for confirming the gas leakage source at present, the misjudgment rate of the confirming method is high, and if the robot falls into local optimum, the detected concentration information is large, and the local area can be misjudged as the leakage source.

Disclosure of Invention

Aiming at the defects of the prior art, the invention provides a method and a system for positioning a gas leakage source of a robot based on a matrix half tensor product, which can intelligently and adaptively select a positioning action through sensor information detected by the robot in real time to ensure that the gas leakage source is quickly and stably positioned.

The technical scheme adopted for solving the technical problems is as follows: the method for positioning the gas leakage source of the robot based on the matrix half tensor product is characterized by comprising the following steps of:

step S1: determining the number of robots participating in gas leakage source positioning, and input variables and output variables of a fuzzy control rule, wherein the input variables are sensor information detected by the robots in real time, and the output variables are positioning behaviors;

step S2: establishing a fuzzy control rule base from input variables to output variables;

step S3: converting the fuzzy control rule into a structural matrix by using a matrix half tensor product theory;

step S4: and selecting corresponding positioning behaviors according to the sensor information detected by the robot in real time.

Further, in step S1, the number of robots is represented as m, where m is equal to or greater than 1; the input variable is denoted x₁、…、x_nN is the type of the sensor, and n is more than or equal to 1; the output variable is denoted y₁、…、y_m，y_mIs the positioning behavior of the mth robot.

Further, the fuzzy control rule in step S2 is:

$R^q:IFx{1}_{}={A_1}^q,\mathrm{...},x_n=A_n^q,{\mathrm{THENy}}_1={B_1}^q,\mathrm{...},y_m=B_m^q$

where Q is 1, … Q, Q is the total number of control rules, x_i、y_jRespectively representing input linguistic variables and output linguistic variables of the fuzzy controller, the values of which are respectivelyx_iAnd y_jAre each k_iAnd s_j，i＝1,…,n、j＝1,…,m。

Further, the process of converting the fuzzy control rule into the structure matrix in step S3 is:

defining the input variables as:

$x_i={\mathrm{\δ}}_{k_i}\[\mathrm{1...}{k}_i\]$

wherein,represents k_i×k_iThe first column of the size unit matrix, i ═ 1, …, n;

defining the output variables as:

$y_j={\mathrm{\δ}}_{s_j}\[\mathrm{1...}{s}_j\]$

wherein,denotes s_j×s_jThe first column of the size unit matrix, j ═ 1, …, m;

the control rule is:

$R^q:{\mathrm{IFx}}_1={\mathrm{\δ}}_{k_1}^{i_1},\mathrm{...},x_n={\mathrm{\δ}}_{k_n}^{i_n},{\mathrm{THENy}}_1={\mathrm{\δ}}_{s_1}^{j_1},\mathrm{...},y_m={\mathrm{\δ}}_{s_m}^{j_m}$

the algebraic expression of the fuzzy rule is:

$\left\{\begin{array}{c}{y}_1=M_1\∝x\\ y_2=M_2\∝x\\ .\\ .\\ .\\ y_m=M_m\∝x\end{array}\right.$

wherein, oc represents a matrix half tensor product operation;

the algebraic form of fuzzy rule can be expressed as

$\left\{\begin{array}{c}{y}_1=M_{1}x\\ y_2=M_{2}x\\ .\\ .\\ .\\ y_m=M_{m}x\end{array}\right.$

Wherein, $x={\∝}_{i=1}^n{x}_i={\mathrm{\δ}}_{k_1}^{i_1}\∝\mathrm{...}\∝{\mathrm{\δ}}_{k_n}^{i_n}={\mathrm{\δ}}_k^i,y_j=y_{s_j}^{j_j},$ represents the ith column of the k × k identity matrix,n is the number of input variables and m is the number of output variables.

Further, in step S4, the linguistic variables are determined according to the sensor information detected by the robot in real time, then expressed in a half tensor algebraic form as the input of the controller, and finally calculated according to the transformed structural matrix to obtain the corresponding positioning behavior.

Preferably, the sensor information includes laser sensor information, concentration sensor information, vision sensor information, and wind speed sensor information.

Preferably, the positioning behaviors include an obstacle avoidance behavior, a gas leakage source positioning behavior, a visual positioning behavior, an upwind search behavior, a path planning behavior, a chemical tendency behavior, and a random search behavior.

The invention also provides a robot gas leakage source positioning system based on the matrix half tensor product, which is characterized by comprising a fuzzy controller and at least one robot system, wherein each robot system comprises a robot and a plurality of sensors, the sensors collect information and send the collected information to the robot, the robot collects the information collected by the sensors and sends the collected information to the fuzzy controller, and the fuzzy controller selects positioning behaviors according to the collected information and sends positioning behavior instructions to the robot for execution.

Preferably, the fuzzy controller comprises a fuzzy control rule base and a structure matrix memory.

Preferably, the sensor information includes laser sensor information, concentration sensor information, vision sensor information, and wind speed sensor information; the robot execution positioning behavior comprises an obstacle avoidance behavior, a gas leakage source positioning behavior, a visual positioning behavior, an upwind searching behavior, a path planning behavior, a chemical trend behavior and a random searching behavior.

The invention provides a novel intelligent positioning method for a gas leakage source of a robot based on a matrix semitensor product by researching the positioning of the gas leakage source and combining the matrix semitensor product method, wherein the robot can adopt a single robot or a plurality of robots. The invention mainly comprises four stages: the first stage is to establish an input-output fuzzy control system between robot sensor information and robot positioning behaviors (namely, input and output of a determination system); the second stage is to determine a control rule base of the fuzzy control system (i.e. control rules between input and output, i.e. different sensor information and the size of the information determine the positioning behavior executed by the robot); the third stage is to determine/calculate a structural matrix of the system controller by utilizing a matrix semitensor product theory and convert complex fuzzy reasoning into a simple algebraic expression; and in the fourth stage, the gas leakage source is intelligently positioned according to the information of the sensor detected by the robot in real time, so that the positioning efficiency and accuracy are ensured.

The invention has the following beneficial effects:

1) the method has the advantages that the positioning is accurate, fuzzy control rules based on actual manual proportioning regulation experience are adopted, the fuzzy rules are subjected to algebraic control, the intelligent control of the humanoid control rules is carried out on the positioning gas leakage source of the robot by fusing the matrix half tensor product theory and the fuzzy control theory and adopting various adjusting modes and control strategies, and the positioning accuracy of the gas leakage source is ensured.

2) The inclusion is strong, firstly, the carrier positioned by the invention can be a plurality of robots or a single robot; and secondly, the positioning behavior is advanced with time, namely the positioning method can conveniently and quickly update the positioning behavior of the robot according to different sensor information and different requirements, namely the positioning method can be degraded into the positioning of single sensor information and can also be expanded into the multi-sensor information fusion positioning.

3) The safety is high, and the robot gas leakage source positioning system constructed by combining the robots replaces professionals or specially trained animals to position toxic and harmful gas leakage sources, so that the phenomenon of life danger is avoided.

4) The flexibility is strong, and the robot gas leakage source positioning system is wider than the area range covered by fixed wireless sensor network positioning, and the positioning flexibility is high.

5) The method has strong practicability, integrates the matrix half tensor product theory and the fuzzy control theory, adopts various adjustment modes and control strategies, and can meet the positioning requirements of gas leakage sources in different environments.

Drawings

FIG. 1 is a block diagram of a positioning system according to the present invention;

FIG. 2 is a flow chart of a positioning method according to the present invention;

FIG. 3 is a flow chart of the present invention for locating a gas leak source.

Detailed Description

In order to clearly explain the technical features of the present invention, the following detailed description of the present invention is provided with reference to the accompanying drawings. The following disclosure provides many different embodiments, or examples, for implementing different features of the invention. To simplify the disclosure of the present invention, the components and arrangements of specific examples are described below. Furthermore, the present invention may repeat reference numerals and/or letters in the various examples. This repetition is for the purpose of simplicity and clarity and does not in itself dictate a relationship between the various embodiments and/or configurations discussed. It should be noted that the components illustrated in the figures are not necessarily drawn to scale. Descriptions of well-known components and processing techniques and procedures are omitted so as to not unnecessarily limit the invention.

As shown in fig. 1, the system for locating the gas leakage source of the robot based on the matrix half tensor product of the present invention includes a fuzzy controller and at least one robot system, each robot system includes a robot and a plurality of sensors. The fuzzy controller comprises a fuzzy control rule base and a structure matrix memory. The robot comprises a robot body, a sensor information acquisition module, a fuzzy controller and a positioning behavior instruction acquisition module, wherein the sensor information acquisition module is used for acquiring information such as laser, concentration, vision and wind speed and sending the acquired information to the robot body, the robot body collects the information acquired by the sensor and then sends the collected information to the fuzzy controller, the fuzzy controller selects the positioning behavior according to the collected information and sends the positioning behavior instruction to the robot body for execution, and the robot body executes corresponding obstacle avoidance behavior, gas leakage source positioning behavior, vision positioning behavior, headwind search behavior, path planning behavior, chemical trend behavior and random search behavior according to the positioning behavior instruction.

As shown in fig. 2, a method for positioning a gas leakage source of a robot based on a matrix half tensor product of the present invention includes the following steps:

step S1: determining the number of robots participating in gas leakage source positioning, and input variables and output variables of a fuzzy control rule, wherein the input variables are sensor information such as laser sensor information, concentration sensor information, visual sensor information and wind speed sensor information detected by the robots in real time, and the output variables are positioning behaviors such as obstacle avoidance behavior, gas leakage source positioning behavior, visual positioning behavior, headwind search behavior, path planning behavior, chemical trend behavior and random search behavior;

step S2: establishing a fuzzy control rule base from input variables to output variables;

step S3: converting the fuzzy control rule into a structural matrix by using a matrix half tensor product theory;

step S4: and the fuzzy controller selects corresponding positioning behaviors according to the sensor information detected by the robot in real time, and sends positioning behavior instructions to the robot for execution, so that the work of positioning the gas leakage source is completed.

The following describes the steps of the method for locating a gas leakage source according to the present invention.

In step S1, the control system configuration is determined, i.e., the number of robots (single robot positioning or multi-robot positioning) and the input variables (sensor information may be detected during robot positioning) and output variables (positioning behavior selected using the configuration matrix based on the detected sensor information) are determined.

The number of the robots is represented as m, m is more than or equal to 1, and when m is 1, a single robot is adopted for positioning; input variablen is the kind of sensor, and when n is equal to 4, it represents that the input variable is 4 kinds of sensor information x₁,…,x₄Respectively laser information, concentration information, visual information and wind information; the output variable is denoted y₁、…、y_m，y_mFor the positioning action of the mth robot, the output variable y is B^sAnd s is 1, …,7, which respectively represents the obstacle avoidance behavior, the gas leakage source positioning behavior, the visual positioning behavior, the headwind search behavior, the path planning behavior, the chemical tropism behavior and the random search behavior.

And step S2, determining the input and output fuzzy control rule base, inputting the sensor information detected by the robot in real time, and outputting the positioning behavior obtained by calculating the corresponding structural matrix.

The fuzzy control rule is as follows:

$R^q:{\mathrm{IFx}}_1={A_1}^q,\mathrm{...},x_n=A_n^q,{\mathrm{THENy}}_1={B_1}^q,\mathrm{...},y_m=B_m^q$

where Q is 1, … Q, Q is the total number of control rules, x_i、y_jRespectively representing input linguistic variables and output linguistic variables of the fuzzy controller, the values of which are respectivelyx_iAnd y_jAre each k_iAnd s_j，i＝1,…,n、j＝1,…,m。

A control rule between input and output is given according to the system structure determined in step S1, and a control rule, i.e., a knowledge base, of a fuzzy control system based on a matrix half tensor product is given according to the characteristics of sensor information and the characteristics of the existing gas leakage source positioning technology, as shown in table 1.

Table 1: control rule for positioning gas leakage source of robot

And step S3, determining a structural matrix of the fuzzy controller by using a matrix semitensor product theory according to the control rule established in the step S2, and converting the complex fuzzy inference into a simple algebraic expression through the structural matrix.

The matrix half tensor product algebra of the control rule is expressed as follows:

the input variables are:

laser information x₁Concentration information x₂Visual information x₃Wind information x₄

The output variables are:

the algebraic expression of the positioning control rules of the gas leakage source of the robot is shown in table 2:

table 2: control rules in the form of algebraic expressions

The input-output matrix half tensor product expression is:

y＝Mx

wherein M is a structural matrix.

Then the structural matrix of the system can be calculated by the control rule table of the robot gas leakage source positioning control system as follows:

M＝₇[111111112211337755465546]

and step S4, selecting corresponding positioning behaviors adaptively and intelligently according to the sensor information detected by the robot in real time, and ensuring that the robot can be quickly and stably positioned to the gas leakage source. Firstly, determining linguistic variables according to sensor information detected by the robot in real time, then expressing the linguistic variables in a semi-tensor product algebraic form to be used as input of a controller, and finally calculating according to a converted structure matrix to obtain corresponding positioning behaviors.

It is assumed that the sensor information detected by the robot in real time is represented as: the distance between the robot and the obstacle is very short, the concentration information is very large, and the visual information is judged to be true when the laser information is measured, namely a suspected gas leakage source exists in a scene, and the wind information is present, namely $x_1={\mathrm{\δ}}_2^1,x_2={\mathrm{\δ}}_3^3,x_3={\mathrm{\δ}}_2^1,x_4={\mathrm{\δ}}_2^1,$ Using the structural matrix, it is possible to determine corresponding regulating actions as

$\begin{array}{c}y=Mx\\ ={\mathrm{\δ}}_7\[111111112211337755465546\]{\mathrm{\δ}}_2^1{\mathrm{\δ}}_3^3{\mathrm{\δ}}_2^1{\mathrm{\δ}}_2^1\\ ={\mathrm{\δ}}_7^2\end{array}$

Namely, the confirmation action of the robot for executing the gas leakage source corresponds to the control rule table.

The specific process of locating a gas leak source using the present invention, as shown in fig. 3, is as follows.

First, discovery of smoke plume

(1.1) judging whether smell information exists, if so, switching to the step (2.1) to track the smoke plume, otherwise, entering the step (1.2);

(1.2) judging whether visual information exists, if so, carrying out visual search and turning to the step (1.1), otherwise, entering the step (1.3);

(1.3) judging whether wind information exists, if so, performing zigzag traversal and transferring to the step (1.1), otherwise, performing spiral traversal and transferring to the step (1.1);

second, smoke plume tracking

(2.1) judging whether the smell information is high, if so, switching to the step (3.1) to confirm the leakage source, and otherwise, entering the step (2.2);

(2.2) judging whether visual information exists, if so, carrying out olfactory and visual fusion search and turning to the step (1.1), otherwise, turning to the step (2.3);

(2.3) judging whether wind information exists, if so, carrying out wind tropism search and transferring to the step (1.1), otherwise, carrying out chemical tropism search and transferring to the step (1.1);

third, leakage source identification

(3.1) judging whether visual information exists, if so, carrying out olfactory and visual fusion search and fusion confirmation, and if not, carrying out concentration method confirmation; thus, the positioning of the gas leakage source is completed.

The foregoing is only a preferred embodiment of the present invention, and it will be apparent to those skilled in the art that various modifications and improvements can be made without departing from the principle of the invention, and such modifications and improvements are also considered to be within the scope of the invention.

Claims (6)

1. The method for positioning the gas leakage source of the robot based on the matrix half tensor product is characterized by comprising the following steps of:

step S1: determining the number of robots participating in gas leakage source positioning, and input variables and output variables of a fuzzy control rule, wherein the input variables are sensor information detected by the robots in real time, and the output variables are positioning behaviors;

step S2: establishing a fuzzy control rule base from input variables to output variables;

step S3: converting the fuzzy control rule into a structural matrix by using a matrix half tensor product theory;

step S4: selecting corresponding positioning behaviors according to sensor information detected by the robot in real time;

in step S1, the number of robots is represented as m, m ≧ 1; the input variable is denoted x₁、…、x_nN is the type of the sensor, and n is more than or equal to 1; the output variable is denoted y₁、…、y_m，y_mPositioning behavior for the mth robot;

the fuzzy control rule in step S2 is:

$R^q:IF{x}_1=A_1^q,...,x_n=A_n^q,THEN{y}_1=B_1^q,...,y_m=B_m^q$

where Q is 1, … Q, Q is the total number of control rules, x_i、y_jRespectively representing input linguistic variables and output linguistic variables of the fuzzy controller, the values of which are respectivelyx_iAnd y_jAre each k_iAnd s_j，i＝1,…,n、j＝1,…,m；

The process of converting the fuzzy control rule into the structural matrix in step S3 is:

defining the input variables as:

$x_i={\mathrm{\δ}}_{k_i}[1...k_i]$

wherein,represents k_i×k_iThe first column of the size unit matrix, i ═ 1, …, n;

defining the output variables as:

$y_j={\mathrm{\δ}}_{s_j}\[1...s_j\]$

wherein,denotes s_j×s_jThe first column of the size unit matrix, j ═ 1, …, m;

the control rule is:

$R^q:IF{x}_1={\mathrm{\δ}}_{k_1}^{i_1},...,x_n={\mathrm{\δ}}_{k_n}^{i_n},THEN{y}_1={\mathrm{\δ}}_{s_1}^{j_1},...,y_m={\mathrm{\δ}}_{s_m}^{j_m}$

the algebraic expression of the fuzzy rule is:

$\left\{\begin{array}{c}{y}_1=M_1\∝x\\ y_2=M_2\∝x\\ \·\\ \·\\ \·\\ y_m=M_m\∝x\end{array}\right.$

wherein, oc represents a matrix half tensor product operation;

the algebraic form of fuzzy rule can be expressed as

$\left\{\begin{array}{c}{y}_1=M_{1}x\\ y_2=M_{2}x\\ \·\\ \·\\ \·\\ y_m=M_{m}x\end{array}\right.$

Wherein, represents the ith column of the k × k identity matrix,n is the number of input variables and m is the number of output variables.

2. The method as claimed in claim 1, wherein in step S4, the linguistic variables are determined according to the sensor information detected by the robot in real time, the linguistic variables are expressed in the algebraic form of the half tensor product and used as the input of the controller, and the corresponding positioning behavior is calculated according to the transformed structural matrix.

3. The method for locating a gas leakage source of a robot according to claim 1 or 2, wherein the sensor information includes laser sensor information, concentration sensor information, vision sensor information and wind speed sensor information.

4. The method for positioning the gas leakage source of the robot based on the matrix half tensor product as claimed in claim 1 or 2, wherein the positioning behaviors comprise an obstacle avoidance behavior, a gas leakage source positioning behavior, a visual positioning behavior, an upwind searching behavior, a path planning behavior, a chemical trend behavior and a random searching behavior.

5. The robot gas leakage source positioning system based on the matrix semi-tensor product is characterized by comprising a fuzzy controller and at least one robot system, wherein each robot system comprises a robot and a plurality of sensors, the sensors collect information and send the collected information to the robot, the robot collects the information collected by the sensors and sends the collected information to the fuzzy controller, and the fuzzy controller selects positioning behaviors according to the collected information and sends positioning behavior instructions to the robot to execute; the fuzzy controller comprises a fuzzy control rule base and a structure matrix memory, wherein the input of the fuzzy control rule base is sensor information detected by the robot in real time, the output of the fuzzy control rule base is positioning behavior obtained by inputting corresponding structure matrix calculation, and the fuzzy control rule is as follows:

$R^q:IF{x}_1=A_1^q,...,x_n=A_n^q,THEN{y}_1=B_1^q,...,y_m=B_m^q$

where Q is 1, … Q, Q is the total number of control rules, x_i、y_jRespectively representing input linguistic variables and output linguistic variables of the fuzzy control rule base, and the values of the input linguistic variables and the output linguistic variables are respectivelyx_iAnd y_jAre each k_iAnd s_jI is 1, …, n and j are 1, …, m and n are the types of sensors, and n is more than or equal to 1; m is the number of the robots, and m is more than or equal to 1.

6. The matrix half-tensor product-based robotic gas leakage source positioning system of claim 5, wherein the sensor information includes laser sensor information, concentration sensor information, vision sensor information, and wind speed sensor information; the robot execution positioning behavior comprises an obstacle avoidance behavior, a gas leakage source positioning behavior, a visual positioning behavior, an upwind searching behavior, a path planning behavior, a chemical trend behavior and a random searching behavior.